Wind of Change: Offshore Wind Farms, Contested Values And

WIND OF CHANGE: OFFSHORE WIND FARMS, CONTESTED VALUES AND

ECOSYSTEM SERVICES

Sarah Catherine Klain

B.A., Reed College, 2003

M.Sc. The University of British Columbia, 2010

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF

THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES

(Resource Management and Environmental Studies)

THE UNIVERSITY OF BRITISH COLUMBIA

(Vancouver)

October 2016

Increasing reliance on renewable energy promises to reduce carbon emissions. Although national-scale polls demonstrate high levels of public support for developing renewable energy, local opposition has led to cancelations of renewable energy projects globally. This dissertation empirically investigates barriers to the siting of offshore wind farms in reference to their perceived risks and benefits; people’s willingness to pay to mitigate environmental risks; values that influence these choices and attitudes; and public deliberation processes used to engage local citizens in decisions about local siting and alternative energy options.

The first study investigates perceptions of offshore wind farm impacts and why risks to some ecosystem services (ES, i.e., benefits from nature to people) may induce greater concern than others. These differences are attributed to the affective ways in which people perceive risk.

Affectively-loaded impacts (e.g., harm to charismatic wildlife, visual intrusion) were assigned greater weight than more easily quantifiable impacts (e.g., displacement of fishing, impact to tourism). This study suggests that government authorities and developers can anticipate and more explicitly address affective dimensions of renewable energy proposals.

The second study quantifies stated preferences for specific attributes of wind farms: effect on marine life, type of ownership, distance from shore, and cost. The strongest preference was for farms that greatly increased biodiversity via artificial reefs at an additional cost of ~$34-

42/month. This research demonstrates substantial willingness to pay for ecologically regenerative renewable energy development. ii

The third study pilots methods on ‘relational values,’ which link people to ecosystems and include associated principles, notions of a good life and virtues. Preliminary results suggest that relational values are distinct from standard methods of measuring ecological worldview and predictive of attitudes towards offshore wind farms.

The fourth study assesses attributes of effective public engagement processes to site renewable energy projects as they played out in three island communities. Amongst the array of criteria for robust analytic deliberative processes, good public engagement may be condensable to two themes: enabling bidirectional deliberative learning and providing community benefit. Attending to these themes may improve relationships among communities, government authorities and developers when deciding if and where to site renewable energy infrastructure.

iii

Preface

Chapters 2, 3, 4 and 5 of this dissertation are distinct manuscripts written with the goal of publication in academic journals. These chapters are meant to stand alone, which results in some repetition across chapters regarding descriptions of the broader research context and methods.

I was responsible for the idea of exploring perceptions of a hypothetical wind farm, creating an animated visualization for the project, as well as the analysis and writing of Chapter 2. I collected data via interviews with the help of a local research assistant. Although I was the lead author of this chapter, my adviser Kai Chan and committee member Terre Satterfield helped me develop the theoretical framing, several research questions and analytical tools to address the questions. I also collaborated with Jim Sinner and Joanne Ellis from Cawthron Institute, who hosted me and others from my UBC lab group in New Zealand. They also provided important background information for my study and valuable feedback on drafts of my interview protocol, data analysis and the resulting manuscript. UBC’s Behavioural Research Ethics Board approved this project (certificate number H14-00842).

I am the lead author of Chapter 3 having conceived of the research questions, selected the methods, conducted the analysis and written the manuscript. Kai Chan contributed with feedback that improved the design, interpretation, analysis and results of this study. I benefited from discussions with Gunilla Oberg about regenerative design. Robin Naidoo and Noah Enelow provided some technical guidance when I built my choice experiment models. Chapter 3 and 4 were approved by UBC’s Behavioural Research Ethics Board (certificate number H15-01325).

Paige Olmsted and I share the role of first author on Chapter 4. Kai Chan, Terre Satterfield,

Paige Olmsted and I collaborated to develop and refine survey questions that Paige Olmsted and

I tested with different populations. Kai Chan and Terre Satterfield recommended statistical approaches, which Paige and I conducted. Paige and I equally shared the writing of the manuscript, which also benefited tremendously from Kai Chan’s and Terre Satterfield’s input.

Chapter 5 was supported by UBC’s Public Scholars Initiative, which seeks to re-imagine the

PhD process via expanding the types of contributions that are recognized as legitimate components of a PhD and dissertation. I oriented this chapter to bring academic literature to bear upon practitioner’s challenges with local rejection of renewable energy systems. As such, this contribution differs from how it would have been structured it if it had been purely an academic exercise (e.g., we selected our study sites based on the partner organization’s experience working with these communities, rather than a more academically rigorous method of selecting sites).

This study was conducted in collaboration with a non-profit organization, Island Institute. Two of my co-authors, Suzanne MacDonald and Nicholas Battista, are staff at this organization. We worked together to identify the main thrust of this project: reflecting on lessons learned from engaging New England island communities with offshore wind. We drafted and distributed a report for public audiences based on our findings [Klain, S., MacDonald, S., & Battista, N.

(2015). Engaging Communities in Offshore Wind (pp. 1–44). Island Institute], which is freely available on Island Institute’s website. I led the analysis and drafting of this manuscript with input from all co-authors. Terre Satterfield and Kai Chan provided critical feedback and guidance on several drafts to help me improve the structure of the manuscript as well as the figures and better relate these lessons learned to academic theories. v

Table of Contents

Abstract ...... ii

Preface ...... iv

Table of Contents ...... vi

List of Tables ...... xii

List of Figures ...... xiii

List of Abbreviations ...... xv

Acknowledgements ...... xvi

Dedication ...... xviii

Chapter 1: Introduction ...... 1

1.1 Dissertation goals ...... 2

1.2 Theoretical underpinnings ...... 3

1.2.1 Social studies of risk ...... 4

1.2.2 Ecosystem services ...... 7

1.2.2.1 Cultural ecosystem services ...... 10

1.2.3 Environmental and relational values ...... 11

1.2.4 Analytic-deliberative processes ...... 14

1.3 Chapter overviews ...... 16

1.4 Summary ...... 20

Chapter 2: Bird killer, industrial intruder or clean energy? Perceiving the risks of offshore wind farms ...... 21

2.1 Introduction ...... 21 vi

2.2 Methods...... 26

2.2.1 Study area ...... 26

2.2.2 Interview sample ...... 29

2.2.3 Interview design ...... 30

2.2.4 Weighting of concerns ...... 31

2.2.5 Risk factor scoring using risk perception theory ...... 33

2.2.6 From scoring risk associated with wind farms to analysis ...... 34

2.2.7 Weighting of benefits ...... 37

2.3 Results ...... 37

2.3.1 Concerns ...... 37

2.3.1.1 Narrative expressions of concern ...... 37

2.3.1.2 Weights assigned to concerns ...... 39

2.3.2 Benefits ...... 41

2.3.2.1 Narrative expressions of benefits and trade-offs ...... 41

2.3.2.2 Weights assigned to benefits ...... 42

2.4 Discussion ...... 43

2.5 Conclusion ...... 47

Chapter 3: Rethinking renewable energy: high willingness to pay for ecologically regenerative offshore wind farms ...... 49

3.1 Introduction ...... 49

3.2 Methods...... 54

3.2.1 Study location ...... 54

3.2.2 Sample characteristics ...... 55 vii

3.2.3 Choice experiment design ...... 57

3.2.4 Econometric analysis of choice experiment data ...... 61

3.3 Results ...... 63

3.3.1 Model results: strong preference for biodiversity benefits ...... 63

3.3.2 Estimates of willingness to pay for offshore wind farm characteristics ...... 65

3.4 Discussion ...... 67

3.4.1 Policy implications ...... 70

3.5 Conclusion ...... 71

Chapter 4: Relational values resonate broadly and differently than intrinsic or instrumental values, or the New Ecological Paradigm ...... 72

4.1 Introduction ...... 72

4.2 Methods...... 77

4.2.1 Survey value statements and sample ...... 78

4.2.1.1 Online survey ...... 80

4.2.1.2 Paper-based survey ...... 81

4.2.1.3 Sampled population characteristics ...... 82

4.2.2 Statistical analysis ...... 83

4.2.2.1 Eigenvalues and scree test ...... 83

4.2.2.2 Factor analysis ...... 84

4.2.2.3 Principal components analysis ...... 85

4.2.2.4 Consistency measure: Cronbach’s alpha ...... 85

4.2.2.5 Correlation testing of environmental values and wind farm attitudes ...... 85

4.3 Results ...... 86 viii

4.3.1 Two distinct factors based on eigenvalues and scree test ...... 86

4.3.2 Factor analysis results: NEP is distinct from relational value ...... 87

4.3.3 Principal components analysis: NEP is distinct from relational values ...... 88

4.3.4 High levels of agreement and consistency with types of environmental value

statements ...... 89

4.3.5 Majority of M-Turk sample have positive attitudes towards wind farms ...... 93

4.3.6 Significant correlations between wind farm attitudes and environmental values ..... 94

4.3.7 Environmental values influence wind farm attitudes at national and state level ...... 96

4.4 Discussion ...... 98

4.4.1 Diverse populations tend to agree with strong relational value statements ...... 98

4.4.2 Relational value responses are distinct from NEP ...... 101

4.4.3 Relational statements can be a single construct and have potential as new index .. 102

4.4.4 Theory implications ...... 104

4.4.5 Policy and practical implications ...... 107

4.4.6 Proposed paths forward ...... 110

4.5 Conclusion ...... 112

Chapter 5: Will communities “open-up” to offshore wind? Lessons learned from New

England islands ...... 113

5.1 Introduction ...... 113

5.1.1 Theorizing public engagement processes ...... 117

5.2 Methods...... 121

5.2.1 Context of study: collaboration with community-based organization ...... 121

5.2.2 Data collection and analysis ...... 123 ix

5.3 Results and discussion ...... 124

5.3.1 Focal island communities and wind farm engagement experiences ...... 125

5.3.1.1 Block Island: the ocean state’s offshore wind farm pioneers ...... 126

5.3.1.2 Martha’s Vineyard: moving forward with a cooperative approach ...... 128

5.3.1.3 Monhegan: confronting deep water and community challenges ...... 130

5.3.2 Bi-directional deliberative learning and community benefits as key to good

engagement ...... 133

5.3.2.1 Defining bi-directional deliberative learning ...... 135

5.3.2.1.1 Readily available and accessible information ...... 136

5.3.2.1.2 Trusted messenger ...... 137

5.3.2.1.3 Bridging organizations ...... 138

5.3.2.1.4 Timing: substantial iterative public engagement before site selection ..... 140

5.3.2.2 Provision of community benefits ...... 143

5.3.2.2.1 Deliberation to determine community benefits ...... 145

5.3.2.2.2 Flexible models for custom tailored benefits ...... 148

5.3.2.3 Relevance to components of public participation in deliberation ...... 149

5.4 Conclusion ...... 151

Chapter 6: Conclusion ...... 153

6.1 Realization of renewable energy research goals and research implications ...... 154

6.2 Limitations ...... 156

6.3 Future research directions ...... 159

6.4 Towards ecologically and socially sustainable energy ...... 160

References ...... 162 x

Appendices ...... 178

Appendix A Golden Bay interview consent form ...... 178

Appendix B Golden Bay Interview request letter ...... 180

Appendix C Golden Bay interview protocol ...... 182

Appendix D Full table of risk components ...... 193

Appendix E Choice experiment consent form ...... 196

Appendix F Choice experiment Mechanical Turk request description ...... 198

Appendix G Choice experiment survey ...... 199

Appendix H Variables in choice experiment ...... 210

Appendix I Factor Analysis by population ...... 211

Appendix J Scree plot ...... 213

Appendix K Graphical PCA results ...... 214

Appendix L M-Turk Cronbach’s alphas ...... 215

Appendix M Variables on wind farm attitudes and indices of environmental value ...... 216

Appendix N Wind farm attitudes ...... 217

Appendix O Distribution of responses to value prompts ...... 218

Appendix P Detailed site descriptions ...... 219

List of Tables

Table 2.1. Common concerns associated with offshore wind farms...... 33

Table 2.2. Psychometric Risk Characteristics (P. Slovic, 1987)...... 35

Table 2.3. Explanation of composite risk factor scoring...... 36

Table 3.1. Survey respondents demographic characteristics compared to census data...... 57

Table 3.2. Description of attributes and levels used in the choice experiment...... 59

Table 3.3. Choice experiment conditional and mixed logit models with WTP estimates (N=400).

...... 64

Table 4.1. Value statements used in surveys...... 79

Table 4.2. Demographic characteristics of our three samples...... 83

Table 4.3. Factor Weights ...... 87

Table 4.4. PCA loadings based on correlation matrix...... 89

Table 4.5. Cronbach’s alpha, mean response and standard deviation of responses across value statements...... 90

Table 4.6. Top six mean responses to environmental value statements across three populations.93

Table 4.7 Linear model results for fixed effects on attitudes towards wind power as anticipated by responses to environmental value statements and demographic characteristics...... 97

Table 5.1. Key differences between New England island study sites and mainland communities relevant to engagement on energy issues...... 126

Table 5.2. Summary of good practices and challenges related to community engagement...... 134

xii

List of Figures

Figure 1.1. Conceptual framework of barriers to scaling up renewable energy...... 17

Figure 2.1. New Zealand electricity generation by source in 2014 (MBIE 2015)...... 27

Figure 2.2. Study Site: Golden Bay, New Zealand...... 29

Figure 2.3. A still image from the animated seascape visualization of an offshore wind farm in

Golden Bay, New Zealand using Google Earth...... 31

Figure 2.4. Relative weighting of offshore wind farm concerns with standard error...... 39

Figure 2.5. Average level of concern according to stakeholders for risks to ES plotted against psychological dimensions of each risk with standard error bars...... 40

Figure 2.6. Perception of relative value of benefits from an offshore wind farm...... 43

Figure 3.1. Wind resource potential for states in study...... 55

Figure 3.2. Example of choice scenario...... 60

Figure 3.3. Willingness to pay (WTP) for offshore wind farm attributes...... 67

Figure 4.1. Graphical results of Factor Analysis...... 88

Figure 4.2. Mean and distribution of responses to relational value prompts and New Ecological

Paradigm Statements...... 91

Figure 4.3. Mean response with standard errors to value prompts across three populations...... 92

Figure 4.4. Attitude toward wind at the national (left) and state level (right)...... 93

Figure 4.5. Expected impact of an offshore wind on going to the coast for recreation...... 94

Figure 4.6 Correlation matrix of attitudes towards wind farms and environmental values...... 95

Figure 4.7. Value-belief norm model (green) with our proposed ways in which relational framings (purple) could influence steps of this pathway (black dashes)...... 106 xiii

Figure 5.1. Normative theory of public participation in decision-making, adapted from Abelson et al. (2003)...... 118

Figure 5.2. Map of focal islands ...... 123

Figure 5.3. A robust approach to developing community benefits...... 147

Figure 5.4. Design and evaluation principles for public participation processes with community benefit outcomes...... 150

xiv

List of Abbreviations

ES Ecosystem services

OWF Offshore wind farm

WTP Willingness to pay

Acknowledgements

I am grateful for Kai Chan’s unwavering confidence in my ability to learn, work through challenges and improve myself as a scientist and citizen. I admire his commitment to his students and unwavering motivation to conduct research that is both academically robust and applied to finding solutions to sustainability issues. His mentorship has profoundly shaped and enhanced my professional trajectory. Also, his Connecting Human and Natural Systems (CHAN’s) lab group provided an intellectual and personal safety net as well as springboard during my academic journey.

Terre Satterfield generously shared her time, wisdom, encouragement and compassion, which was invaluable as my dissertation directions evolved. I am indebted to my committee members, including Terre Satterfield, Hisham Zereffi and Scott Harrison, for their support and remarkable patience with me.

My research in New Zealand was made possible due to Jim Sinner and Joanne Ellis at Cawthron

Institute. Our collaboration was funded by the Ministry of Business, Innovation and Employment

(MBIE) (contract MAUX1208). Evan Jones provided essential animation assistance. I also thank the following research assistants for their help in conducting and transcribing interviews:

Ruaridh Davies, Jakob Öberg, Allison Thompson, Calum Watt and Adrian Semmelink.

I also appreciate the financial support from the Vanier Fellowship. UBC’s Public Scholar

Initiative and The Biodiversity Research: Integrative Training & Education (BRITE) Natural xvi

Sciences and Engineering Research Council of Canada (NSERC) Collaborative Research and

Training Experience Program (CREATE) program enabled me to collaborate with the non-profit organization Island Institute. Suzanne MacDonald, Brooks Winner, Harry Podolsky, Rebecca

Clark Uchenna at Island Institute made me feel welcome and shared their considerable knowledge and experiences that shaped our work. Additional support for my dissertation came from the Social Sciences and Humanities Research Council of Canada (SSHRC) grant F12-

04439 Environmental meanings and ecosystem services: the social risks of ecological change and the Gordon and Betty Moore Foundation.

I am also grateful to my family and friends who provided emotional support during my PhD process. I appreciate my mother’s boundless energy and ability to cut to the chase. My father’s love of the sea, sailing and wind has rubbed off on me. My accomplished sister and her lovely daughter inspire me to do my best.

xvii

Dedication

To Josephine Ellen Kellner-Klain, with love. You are the future and a considerable part of why I devote myself to addressing sustainability challenges.

xviii

Chapter 1: Introduction

“When we choose the kind of nature we will live with, we are also choosing the kind of human beings we will be. We shape the world, and it shapes us in return. We are the creator and the created, the maker and the made.”

~J.B. MacKinnon

Securing sustainable energy is among humanity’s most urgent problems, particularly in the context of climate change (Yergin, 2011). Energy choices involve trade-offs replete with environmental, economic and social consequences. Over 1,300 scientists around the world have prioritized the following human impacts as key global concerns: climate disruption, extinctions, loss of diverse ecosystems, pollution, and human population growth in conjunction with high levels of consumption (Barnosky et al., 2014). These concerns are directly and indirectly linked to the production and consumption of energy for human use.

The scientific consensus on the need to decrease greenhouse gas emissions has coalesced.

Part of mitigating climate change involves decarbonization—reducing the carbon intensity of energy (IPCC, 2014). Rapidly scaling up low carbon electricity production to replace energy from fossil fuels plays a crucial role in decarbonization goals set by countries around the world during COP21 (Bagheri and Del Amo, 2016) and can help achieve the United Nation’s Sustainable Development goals relevant to energy (Angelou et al., 2013; UN, 2015). Renewable energy development is part of the requisite energy transition to mitigate climate change. 1

Numerous studies demonstrate broad public support for renewable energy development in general (Krohn and Damborg, 1999; Krosnick and MacInnis, 2013; Toke, 2002; G.

Walker, 1995; Wüstenhagen et al., 2007). Despite this widespread support at national levels, when it comes to siting specific new energy technologies, many vociferously debate what constitutes clean and locally desirable energy systems (J. Barry et al., 2008;

Devine-Wright, 2011; Devine-Wright et al., 2011; Roberts et al., 2013; Warren et al.,

2005). In modern democratic societies, local opposition to particular renewable energy infrastructure poses a challenge to rapid decarbonization because it can shape if and how energy infrastructure is built (Ansolabehere and Konisky, 2014; Devine-Wright, 2005;

Devine-Wright et al., 2011; Roberts et al., 2013). Fierce local resistance to proposed energy infrastructure has stalled or stopped some energy developments, including cases where federal and regional approval has been granted. For example, the Cape Wind offshore wind farm in the US instigated vigorous local opposition resulting in multiple lawsuits against the project proponents despite governmental approval (Firestone and

Kempton, 2007; Shellenberger and T. Nordhaus, 2009). Within this context of global climate change coupled with significant local opposition to renewable energy projects, this dissertation seeks to identify, characterize and anticipate perceptions of risks, benefits and trade-offs associated with the development of offshore wind farms.

1.1 Dissertation goals

The purpose of this dissertation is to provide insight on 1) the source and nature of resistance to and conflicts surrounding some renewable energy development— specifically offshore wind farms; and 2) identify novel approaches and opportunities for 2

working through such conflicts. It is my hope that this research will inform options for transitioning to low carbon electricity sources in a socially and environmentally responsible manner.

More specifically, I aim to better understand and address how people perceive this renewable energy technology, what trade-offs they are willing to make in light of its environmental impacts and costs and how decision processes about this technology can open up rather than close down public involvement in decisions about this technology.

My research also addresses the critique that energy research has downplayed “the role of choice and the human dimensions of energy use and environmental change” (Sovacool,

2014, p. 1) and suffered from a lack of “human-centered research methods” (e.g., field research, interviews, focus groups, surveys) (Sovacool, 2014, p. 2).

1.2 Theoretical underpinnings

Social theories of perceived risk are central to these research goals, as are emerging characterizations of ecosystem services (ES) and environmental values as pertains to the environment and energy. Also pertinent to this work are theories of public engagement in policy decisions, particularly regarding the design of analytic-deliberative processes involving public groups in decision making and siting. Topically, the focus here on offshore wind farms is also, by definition, a testing ground for new ways of applying, hybridizing and contributing to these theories and fields of inquiry.

1.2.1 Social studies of risk Risk perception is central to the choices people make regarding both energy use and their support or opposition to the development of new sources of energy. Public risk perception can profoundly push, constrain or impede action to address specific risks

(Leiserowitz, 2006). Risk perceptions are critical elements of the social and political context in which policy develops and is implemented. Perceptions, rather than technical knowledge per se, drive human behavior (Bennett, 2016; Leiserowitz, 2006).

Understanding perceptions sheds light on what worries and motivates people.

The field of risk perception has been used broadly to understand why people accept or reject new technologies, design communication and education efforts and create robust risk management strategies (Haidt, 2001; Satterfield et al., 2009; P. Slovic, 1987;

Wilsdon and Willis, 2004). The risk perception literature has delved into how people integrate affective (“system 1”) and deliberative (“system 2”) cognition when forming risk judgments (Epstein, 1994; Finucane et al., 2000a; Loewenstein et al., 2001; Sloman,

1996; P. Slovic, 2010).

The psychometric paradigm in perceived risk research is foundational to this newer ‘two- system’ thinking in that it first demonstrated the intuitive nature of risk judgments.

Central to this is empirical work on how perceived risk is both predictable and quantifiable based on a limited set of often intuitive and affective factors, including how well a risk is understood, how much it invokes dread, whether or not a risk is seen as controllable and how many people are thought to be exposed to it. Risk perception studies have proliferated and now attend to more affective and social considerations, 4

including race, gender, vulnerability, and trust (Bord and O'Connor, 1997; Finucane et al., 2000b; Irwin and Wynne, 2004; Satterfield et al., 2004). Nonetheless, these psychometric dimensions continue to explain much of the variance in perceived risk for both new and familiar technologies (Helgeson et al., 2012; Satterfield et al., 2009).

The psychometric paradigm was a precursor to five categories of influences on the formation of risk perception at the individual level including cognitive, sub-conscious, affective, socio-cultural, individual factors (Helgeson et al., 2012). Cognitive factors include expected utility and rational estimations of likelihood and impact. Subconscious drivers include cognitive heuristics (rules of thumb) that can lead to substantial and persistent biases (e.g., misunderstandings of probabilistic processes) (Kahneman, 2011;

Leiserowitz, 2006; Tversky and Kahneman, 1974). Affective factors, including like

(positive valence), dislike (negative valence), fear, anxiety and worry, tend to direct how we process information and make judgments about risks (Finucane et al., 2000a;

Loewenstein et al., 2001). Socio-cultural factors that influence risk perceptions include social organization (hierarchical versus egalitarian) and social relations or cohesion (high conformation to norms versus loose conformation) as well as broadly shared values and beliefs constituting worldviews (Douglas and Wildavsky, 1983). Risk perception can be linked to commitments to cultural and political groups, as explained with the cultural theory of risk (Douglas and Wildavsky, 1983; Kahan, 2015; Kahan et al., 2012). Lastly, the social amplification of risk framework demonstrates how communications with different qualities and from different communication sources (media, NGOs, etc.) can amplify or attenuate risk perception (Kasperson et al., 1988; Pidgeon et al., 2003).

Individual factors also play a role in risk perception formation, notably that people with low levels of self-efficacy, which is an individual’s perception of his/her capacity to instigate change in his/her life, experience higher levels of perceived personal risk.

Additionally, direct experience of a risk also strongly influences risk perception

(Helgeson et al., 2012).

Risk perception research has thus far largely focused on personal health and safety concerns, but some opposition to energy infrastructure stems from environmental considerations (e.g., polluted or destroyed habitat) (Ansolabehere and Konisky, 2014), particularly the reduction of ecosystem services (ES), defined as benefits from the environment to people (e.g., fisheries, freshwater) (Entrekin et al., 2011). In contrast to many of the foci of risk research (e.g., radiation, natural disasters), which have direct consequences for human health and safety, risks to ES tend to have more indirect impacts to people.

I apply risk perception theories in a new context: perceptions of the ecological risks posed by the development of a renewable energy technology. Understanding these perceptions of risks to ES could help design mitigation strategies for local environmental impacts and potentially garner greater public support for transitioning away from fossil fuels. I seek to better understand intuitive risk judgments and perceptions of benefits associated with renewable energy infrastructure. My research in Chapter 2 tests the extent to which the psychometric risk paradigm can be extended to and help explain the magnitude of locally perceived risks to ES.

I first address this set of risk perception research challenges in Chapter 2 with the question: Can the psychometric risk paradigm be extended beyond human health and safety concerns to less direct risks mediated by the environment—e.g., can it predict perception of ecological risk associated with new energy infrastructure? How do people perceive environmental risk associated with a new technology? And do such applications of the psychometric risk paradigm helps anticipate the salience of ES impacts to stakeholders in relation to a new renewable energy technology? In sum, the aim of that chapter is to use an illustrative case study to provide a proof of concept for bringing together ES and risk perception literature.

1.2.2 Ecosystem services The concept of ES emerged in the early 1980s to characterize the subset of ecological functions that are valuable to people but not always captured by conventional cost-benefit approaches (P. R. Ehrlich and A. H. Ehrlich, 1982; P. R. Ehrlich and Mooney, 1983;

Kremen, 2005). ES became more mainstream after more than 1000 scientists around the world collaborated to write the Millennium Ecosystem Assessment (MA, 2003), which launched the concept on a global stage (Abson et al., 2014; Gómez-Baggethun et al.,

2010). ES as a research field seeks to identify, quantify and value the benefits that nature provides to people (G. C. Daily, 1997; MA, 2003). Considerable effort has been invested into developing strategies to integrate ES into natural resource decision-making at multiple scales (G. Daily and Matson, 2008; G. C. Daily et al., 2009). The ES framework has become a common structure with which to identify and categorize the benefits that nature provides to people in ways designed to inform decision-making (Guerry et al.,

2015; Ruckelshaus et al., 2013; Tallis and Polasky, 2011). ES research often includes the estimation of trade-offs across multiple ES depending on the location and type of development (Kareiva et al., 2011). I use ES to categorize environmental impacts because this field of research highlights the connections between environmental changes and changes in benefits that people derive from ecosystems.

Risk perception research has not yet been substantially integrated into the ES research agenda. This integration is important because of increasing recognition that perceptions of and decisions relevant to ES, similar to risk, are largely about non-material values and considerations. That is, ES are most salient when there is perception of real or potential harm or loss rather than just a static provision of a service.

Understanding risk perceptions of new technologies and how they impact ES is crucial because, as previously noted, perceptions drive human behavior (Bennett, 2016;

Leiserowitz, 2006). When a risk is already controversial, new information about scientifically assessed risks of a technology does not easily change preconceived perceptions and biases (Leiserowitz, 2006; Satterfield et al., 2009). If a risk, however, is not widely known and not (yet) controversial, new information can shift perceptions of risk (Allum et al., 2008; Satterfield et al., 2012).

ES assessments have the potential to provide new information that can clarify trade-offs associated with management options and inform decision-making. They generally focus on the consequences of a natural resource management decision on the benefits that

people derive from ecosystems. In practice, these consequences are often only a small part of what drives stakeholder support, consternation, and/or rejection (Gregory et al.,

2012; Spash, 2008a). Part of my motivation for Chapter 2 is based on the premise that researchers and people conducting ES assessments and associated decision-making processes could likely better anticipate (and potentially change) levels of support for a project or policy if they had a better understanding of some psychological dimensions of

ES perceptions.

This dissertation also explores perceptions of ES change and level of support for particular changes. Many critique ES valuations for their limited uptake in real-world contexts (Förster et al., 2015; Honey-Rosés and Pendleton, 2013; Martínez-Harms et al.,

2015), perhaps because they are not sufficiently specific as to what people would pay via realistic payment vehicles for ES protection or restoration. Chapter 3 aims to estimate how much people would be willing to pay for a feasible ecologically beneficial artificial reef in conjunction with an offshore wind farm.

The widespread uptake of the concept of ES in government (SAB, 2009), non- governmental organizations (Tallis et al., 2010), academia (Seppelt et al., 2011), global financial institutions (Naber et al., 2009), and to a growing extent corporations (Tercek and J. S. Adams, 2013), has left many uncomfortable with the way ES assessments tend to embrace an anthropocentric, often individualistic and consumer-oriented worldview, replete with the language of markets, producers, consumers and dollar values attached to nature rather than emphasis on nature’s intrinsic value (W. M. Adams, 2014; Spash,

2008b). Research on cultural ecosystem services, defined as “ecosystems' contributions to the non-material benefits (e.g., capabilities and experiences) that arise from human– ecosystem relationships” (Chan et al., 2012b), has critiqued this market value orientation while also attempting to broaden the types of values integrated into ES assessments

(Chan et al., 2012a; Daniel et al., 2012; Klain and Chan, 2012).

1.2.2.1 Cultural ecosystem services

As a field, ES has tended to emphasize the instrumental value of nature — nature is valuable because it is useful to people. Numerous ES studies have estimated the instrumental value of provisioning, supporting and regulating ES, but instrumental and monetized value falls short when identifying, assessing and characterizing cultural ES

(Chan et al., 2012b; Daniel et al., 2012). Instrumental values are substitutable, but cultural values are often not (Chan et al., 2011; 2012b). Quantified and/or monetized ES data often omit intangible values, including connectedness and belonging to a community

(both human and non-human), sense of place and other culturally and psychologically mediated relationships between people and ecosystems (Russell et al., 2013). This led researchers from diverse fields, not just ecology and economics which dominated earlier

ES studies, to design and test methods aimed at enabling social, cultural and intangible values to play a more prominent role in ES assessments and decision-making

(Chan et al., 2012b; 2012a; Daniel et al., 2012; Gould et al., 2014; Klain and Chan, 2012;

Martín-López et al., 2009; 2012; Milcu et al., 2013; 2014; Plieninger et al., 2013; Sherren et al., 2010). One new frontier along this cultural ES research trajectory is testing

relational value framing, which may motivate pro-environmental behavior (Chan et al.,

2016).

1.2.3 Environmental and relational values

People concerned about climate change, the biodiversity crisis and other ecologically detrimental anthropogenic impacts often propose changing human values as a means to achieving more sustainable behavior and policies (Dietz et al., 2005; Nichols, 2014).

Values can be defined as assigned values (degree of goodness, worth, importance or meaning that people put on an object) or held values (underlying ideals)(Brown, 1984).

Identifying, characterizing and quantifying the “value” of nature underpins ES research.

The original architects of the field of ES explained their research as an attempt to highlight the value of nature in ways that were previously overlooked with the assumption that this information would push decision-making towards more nature- friendly outcomes (G. C. Daily, 1997; MA, 2003; Spash, 2008b). This “value” of nature in ES literature has often been summed up in monetary value (Costanza et al., 1998; Karp et al., 2013), which has limitations and, in some contexts, may not benefit biodiversity or conservation (W. M. Adams, 2014).

One of the perils of ES approaches that emphasize monetary valuation is that money and appeals to financial benefit and self-interest reinforce extrinsic values, which are associated with the pursuit of prestige, power, image and status. Psychological research has shown that reinforcement of extrinsic values can suppress intrinsic values, which are

linked to concern for others and the environment, kindness, understanding, appreciation, tolerance and protection of people and nature (Blackmore et al., 2013). Intrinsic motivations for conservation—protecting nature for its own sake—has driven many conservation biologists and conservation efforts, but such environmental value framing is critiqued as being overly narrow (Marvier and Wong, 2012), lacking appeal to diverse audiences and deaf to the needs of many people, particularly poor people (Kareiva et al.,

2012).

Relational-value framing might be more broadly appealing and motivating to pro- environmental behavior than instrumental and intrinsic value framing. Relational values include “eudaimonic” values, defined as those related to living a good life, justice, reciprocity, care and virtue (Jax et al., 2013; Muraca, 2011; Ryan and Deci, 2001; Ryff and Singer, 2008). Interactions with and responsibilities to humans, non-humans, landscapes and ecosystems give rise to relational values (Chan et al., 2016). Research on relational values in the context of social-ecological interactions has been lacking. Chapter

5 uses quantitative methods to test the application of social-ecological relational statements as preliminary steps towards further testing if such value framing can enhance connection to the natural world and pro-environmental behavior and policies.

Diverse and often conflicting environmental values come into play when considering if and where to build renewable energy infrastructure. The “green-on-green” debate about wind farms characterizes conflict related to the extent to which environmentally minded stakeholders prioritize global environmental concern (i.e., climate change) versus local

environmental concern (e.g., bird strikes from wind turbines, aesthetic degradation of landscape) (Warren et al., 2005).

More explicit consideration of relational values, broadly conceived, may be key to addressing renewable energy and other sustainability issues. Activating relational values focused on concern for and protection of people and the environment could help change individual and collective behavior, policies and ultimately society’s relationship to nature. The types of relationships with ecosystems that we choose may “change everything,” in the words of Klein (2014), who advocates transitioning from extractive to regenerative systems:

Extractivism is a nonreciprocal, dominance-based relationship with the earth, one purely of taking. It is the opposite of stewardship, which involves taking but also taking care that regeneration and future life continue. Extractivism is the mentality of the mountaintop remover and the old-growth clear-cutter. It is the reduction of life into objects for the use of others, giving them no integrity or value of their own— turning living complex ecosystems into “natural resources,”... In an extractivist economy, the interconnections among these various objectified components of life are ignored; the consequences of severing them are of no concern (Klein, 2014, p. 169).

In a regenerative system, links between components of ecosystems are recognized.

Regenerative systems increase diversity, require little external inputs and produce virtually no waste. Such systems promise restoration, renewal and revitalization (Lyle,

1996; McDonough and Braungart, 2002). Regenerative technology is increasingly common in medical sciences but not yet prominent in conservation efforts. Regenerative design concepts, which can enhance biodiversity while providing for human needs, have

not yet been applied to offshore wind farms. Chapter 3 assesses the extent to which ecologically regenerative wind farm characteristics might affect preferences and willingness to pay for this technology.

Relational values in the context of energy transitions raise many questions addressed in

Chapter 5, such as what improves or erodes the quality of the relationships between wind farm developers, government authorities and local communities? What role should community benefits from developers play in the decision process? What environmental mitigation efforts should be taken to offset the local environmental impacts? Such relational value questions could prime analytic-deliberative processes to increase the likelihood of reaching legitimate outcomes when it comes to considering and siting renewable energy infrastructure.

1.2.4 Analytic-deliberative processes

Robust public engagement strategies may help to assuage renewable energy controversies. Accordingly, this dissertation draws upon literatures focused on analytic- deliberative processes of engagement that have the potential, in the words of Stirling

(2008) to “open-up” rather than “close down” discussions about new technologies and innovations.

Abelson et al. (2003) and Ryfe (2005) review the normative theory of public participation in decision-making. Abelson et al. (2003) operationalizes this theory into pragmatic evaluation principles with explicit recognition of the role of power in deliberative

processes. These reviews emphasize how there is no simple formula for an optimal public engagement process, but four key issues deserve attention: 1) representation; 2) procedural rules; 3) information employed in the process and 4) the outcomes including decisions resulting from the process. Representation determines who represents the

“public, ” which poses challenges. Namely, legitimate and fair processes provide meaningful opportunities for learning and recognize diverse perspectives, so consequently tend to be time-intensive and relatively exclusive processes in which it is only feasible to involve a small number of people. Also, citizens are more likely to get involved if they fear losing something they value, which further complicates fair representation (Abelson et al., 2003). Situations can arise when a majority of people support or are neutral towards a proposal, but they are a “silent majority” because they opt not to get involved with the decision process (Stephenson and Lawson, 2013).

Abelson et al. (2003) documents how procedural rules can help manage this potential self-selection of who gets involved. Choices about information are crucial, specifically what information is selected then how it is presented and interpreted. Finally, not just the process leading to the decision, but also the outcome (the decision) needs to be associated with legitimacy and accountability (Abelson et al., 2003).

These evaluation criteria were developed in the heath sector, but much of them apply to a wide array of contexts, including decisions involving communities about renewable energy. The length and complexity of analytic-deliberative process features deemed important to reach legitimate conclusions is likely overwhelming to practitioners. Based on field work in three island communities that have considered offshore wind farm

development, I derived a shorter, more practitioner-friendly list of key design features of both the decision process and an outcome, specifically bi-directional deliberative learning and the provision of community benefits.

1.3 Chapter overviews

Broadly stated and referenced above, I situate my dissertation within a broader context of the major barriers to scaling up renewable energy. Figure 1 below depicts my conceptual framework of barriers to scaling up renewable energy, including national and regional- scale obstacles to the rapid expansion of renewables. I do not, however, incorporate these national and regional barriers pertaining to finance and policy in this dissertation. Instead,

I focus on facets of public opposition, which tend to operate at local and regional scales.

The following chapter overviews focus on different sources of public opposition to offshore wind farm development.

Na9onal & Regional Regional & Local Government Financial & Policy Public Opposi9on

Insufficient long-term Structure of planning to implement Concerns Flawed electricity markets Value renewable energy targets about engagement à dominant players à Uncertainty orienta9ons suppress consequences processes newcomers Ineffec9ve communica9on between High expense to government and Local Social Financial create new grid regulatory bodies environment infrastructure where à Confusion for renewable resource developers and delays is most abundant Shortage of experienced Provisioning Cultural Biodiversity Community Lack of staff in government and ecosystem ecosystem benefits do not realized regulatory agencies services services jus9fy burden WTP à Delays and uncertainty

Figure 1.1. Conceptual framework of barriers to scaling up renewable energy. National and regional financial, governance and policy issues impede the proliferation of renewable energy development globally (WEF, 2011). Public opposition to renewables, the main topic of this dissertation, operates at local and regional scales. Each dissertation chapter focuses on different elements of public opposition (topics in blue boxes). Source for gray box: Economic and Government & Policy: WEF. (2011). Scaling up renewables (pp. 1–48). World Economic Forum. Geneva.

Chapter 2, Bird Killer, Industrial Intruder or Clean Energy? Perceiving the Risks of

Offshore Wind Farms brings together ES research with risk perception theory, specifically the psychometric risk paradigm. This research touches upon concerns about consequences particularly as related to the local environment as well as value orientations

(see Figure 1.1). The study context is a hypothetical wind farm in a location with excellent wind quality near an area of high bird diversity and abundance. This research, which uses an animated wind farm seascape visualization, addresses the questions: when considering offshore wind farms, what risks and benefits do people perceive? What is the relative magnitude of how people perceive risks to ES? It tests the hypothesis that features of the psychometric risk paradigm predict relative levels of risk associated with impacts from an offshore wind farm. The results suggest that attributes of this risk paradigm do indeed apply to concerns about ES. Also, this kind of anticipation of risk perceptions can contribute to technology designs that better reflect citizens’ risk perceptions.

Technology design plays a crucial role in Chapter 3, Rethinking renewable energy: High willingness to pay for ecologically regenerative offshore wind farms. This chapter delves into concern about consequences, the local environment and lack of realized willingness to pay within the conceptual framework. This choice experiment addresses: Is there latent willingness to pay for ecologically regenerative renewable energy infrastructure? More specifically, what if offshore wind farms provide high quality marine habitat via artificial reefs? I test the hypothesis that people are willing to pay more than current utility rates for an offshore wind farm that provides marine biodiversity benefits. This study also

18 assesses willingness to pay to reduce the visual impact of an offshore wind farm (i.e., increase the distance from shore) and preferences for ownership type (i.e., state, municipal, private or cooperative).

Chapter 3 focuses on assessing preferences and willingness to pay while Chapter 4,

Relational Values Resonate Broadly and Differently than Intrinsic or Instrumental

Values, or the New Ecological Paradigm (NEP), explores the extent to which relational value statements resonate across three distinct populations. This exploratory research characterizes the extent to which relational values resonate differently than purely instrumental or intrinsic values. Also, it tests for correlation between the strength of environmental values measured and attitudes towards wind farms. It explores value orientations associated with opposition and support of wind power as well as concerns about consequences to the local environment as depicted in the conceptual framework

(Figure 1.1).

Chapters 2 through 4 used a hypothetical wind farm. In contrast, Chapter 5, Will communities “open-up” to offshore wind? Lessons learned from New England Islands focuses on three New England islands near proposed wind farms. This chapter addresses public opposition arising from concerns about consequences, value orientations, and flawed engagement processes. All of the topics linked to public opposition in the conceptual diagram in Figure 1.1 connect to Chapter 5. This chapter streamlines best practices and design principles for analytic-deliberative processes that can improve the quality of the relationships between wind farm developers, government authorities and

19 local communities. It explores features of decision processes that built or eroded trust, including explicit consideration of community benefits. This has implications for a range of development proposals where one scale or group of interest imposes on another.

1.4 Summary

In the aggregate, this dissertation explores theory and empirical data relevant to meeting the challenge of climate change while promoting rigorous decision processes, encouraging reflection on social-ecological relational values and protecting biodiversity.

This background of climate change and energy infrastructure controversy in general and the offshore renewable energy frontier in particular provides a novel context for the application of and contribution to the fields of social studies of risk, ES, environmental and relational values and analytic-deliberative process design. The conclusion highlights the main findings of my research, recommendations for future studies and implications for practitioners. This research helps reconcile renewable energy development and biodiversity conservation. It aims to clarify psychological attributes that influence perceptions of ES change, assess support for ecologically regenerative renewable energy, explore relational values and contributes to improving public participation in decisions about renewable energy. Together these insights provide collaborative and proactive approaches to creating new energy systems that are more conducive to long term prosperity for human and non-human life.

20 Chapter 2: Bird killer, industrial intruder or clean energy? Perceiving the risks of offshore wind farms

Sarah C. Klain, Terre Satterfield, Jim Sinner, Joanne I. Ellis, Kai M.A. Chan

2.1 Introduction

A central strategy for climate change mitigation entails the replacement of existing sources of energy with low carbon renewable energy (Hoffert, 2002; IPCC, 2011). The speed and scale at which renewables are deployed and fossil fuels phased out will have significant consequences on the world’s climate trajectory (Moss et al., 2010; W. D.

Nordhaus, 2013). Local opposition to renewable energy development is a major challenge to transitioning to low carbon technologies since it can shape if and how energy infrastructure is built (Ansolabehere and Konisky, 2014; Devine-Wright, 2005; Roberts et al., 2013). Such opposition can be a function of numerous factors, including but not limited to actual and perceived economic costs, inequitable distribution of costs and benefits, unfair siting processes and unacceptable risks associated with the development, such as the risk of environmental impacts (Bell et al., 2005; Devine-Wright, 2005;

Roberts et al., 2013; Wolsink, 2000).

While recognizing the numerous facets of the social acceptance of new technologies, we focus here on risk perception, which has been widely used to understand some predictable patterns, logics and mental models that underpin evaluations of new technologies (P. Slovic, 1999). In particular, this literature has documented the role of

21 what is known as dual processing theories of cognition: how people integrate affective

(“risk as feelings”) and deliberative (“risk as analysis”) cognition when forming risk judgments (Finucane et al., 2000a; Loewenstein et al., 2001; P. Slovic, 2010; P. Slovic and Peters, 2006).

Qualitative understandings—meanings—influence people’s perceptions of risk, in addition to, and perhaps even more than, quantitative information (P. Slovic, 2010). In this sense, studies of risk perceptions have demonstrated how perceived risk is both predictable and quantifiable based on a limited set of often intuitive and affective factors, including the extent to which a risk is understood, who is exposed, and whether or not the object in question invokes dread, which can be defined as extreme fear or anxiety regarding future events (P. Slovic, 1987)(for a full list of factors, see Table 2.1). This research, typically conducted with expressed preference surveys, has sought to explain why and how people evaluate a hazard according to various psychometric rating scales

(e.g., severity of consequences, novelty). Risk research has evolved to focus more on affective responses (Loewenstein et al., 2001; P. Slovic, 2010; S. Slovic and P. Slovic,

2010), but we use the psychometric risk paradigm because it helps explain why people have affective responses to particular risks. The psychometric risk paradigm theorizes that perceived risk is both predictable and quantifiable based on the extent to which the risk is known to science and dreaded/affectively loaded (Slovic, 2000).

Risk perception studies have also generally focused on risks of direct harm to personal health with less attention paid to environmental risks. We see an opportunity to integrate

22 ecosystem services approaches into the risk literature. Scientists and practitioners have used the ecosystem services (ES) framework to identify, quantify and often estimate a monetary value for the human consequences of environmental impacts. However, ES as a field has focused primarily on impacts as quantified biophysically and often translated into monetary terms to highlight benefits from nature that could be lost depending on development choices (G. C. Daily, 1997; Kareiva et al., 2011; Nelson et al., 2009) (e.g., a specified tract of forest in a watershed provides x amount of clean water worth $y). There has been little attention to understanding how some services and benefits at risk from infrastructure development might be cause for greater public concern than others based on the affective and intuitive ways by which people perceive risk.

Thus far, risk perception theory has been tested primarily in the context of direct risks to human health and safety, rather than risks to one’s broader sense of well-being as experienced via loss or degradation of ES. This paper addresses the broad question: do the same logics by which some personal risks loom larger than others also apply to the context of perceiving risks to ES?

Our research applies risk theory and methods in a new context: perceptions of the risks posed by the development of an offshore wind farm as mediated by the environment.

That is, people remain those judging the risks, but instead of evaluating risk to human health or even environmental health (e.g., air quality), we instead attempt to measure the relative level of concern associated with risk to various ESs.

23 For instance, we assess the relative magnitude of concern associated with the risk that an offshore wind farm would pose to birds, which tends to be a prominent concern based on public surveys (Firestone et al., 2009; Warren et al., 2005), as compared to other ecosystem services (ES). We hypothesize that the relative weighting of various risks to

ES follows the logic of the psychometric theories of risk, which posits that the relative weight of risks will follow the degree to which an impact is affectively loaded and/or dreaded and unknown to science (Slovic, 2000).

Results from early studies based on the psychometric paradigm are now interpreted as derivative of the affect heuristic (P. Slovic et al., 2007). The affect heuristic explains how feelings or emotions often precede and drive judgments of risk and benefit. Instead of judging potential outcomes impartially, people tend to judge risks based on immediate emotional reactions. Non-experts generally perceive an inverse relationship between risk and benefit; high-risk activities or technologies are associated with low benefits and vice versa. If people like or, in other words, attach positive affect to an activity or technology, they tend to see associated risks as low and benefits as high. If they dislike it, they will associate it with high risk and low benefits (Finucane et al., 2000a). Feelings of dread are now seen as predictors of a high level of perceived risk because dread is an affectively loaded quality.

Such affective aspects of risk perception are likely key for understanding why some proposed energy projects elicit highly charged resistance. Understanding these risk perceptions and what drives them is particularly important because renewable energy

24 infrastructure and risks associated with them are likely to be increasingly salient to people as such technologies become more widely known and prominent in inhabited landscapes.

In this article, we thus test theories of risk as applicable to the changes in ES potentially introduced by an offshore wind farm. Our investigation focuses on ES concerns associated with both tangible (e.g., commercial fisheries) and intangible services (e.g., aesthetic value as assessed by perception of negative visual impact). Our illustrative case study provides a proof of concept for integrating risk perception and ES literatures. We seek to advance the integration of risk perception theory and method into ES assessment and research agendas and inform mitigation strategies for local environmental and social impacts of renewable energy. Another aim is to contribute to understanding of public support or rejection for energy transition options. In so doing, we address three research questions:

1. On a relative scale, what are study participants most concerned about when it comes to the development of an offshore wind farm? 2. Do psychometric risk dimensions and the associated affect heuristic predict how study participants weight potential consequences of the risk from wind farms to the provision of ES? 3. On a relative scale, what do study participants perceive as important benefits associated with an offshore wind farm?

25 2.2 Methods

We used semi-structured interviews to ask two overarching questions: What risks associated with a hypothetical offshore wind farm are most salient to people who live near potential wind farm sites? What benefits are most salient?

The hypothetical wind farm site is physically well suited for the technology, but no wind farm proposal currently exists for the site. Participants’ perceptions were not influenced by local campaigns for or against an offshore wind farms since such campaigns were nonexistent. The interviewer provided brief background materials using neutral language about energy, renewable energy, and offshore wind farms, followed by a visualization of an offshore wind farm in a location familiar to participants. Participants were asked about their perceived impacts to ES and opinions on offshore wind farms and then asked to assign weights to a variety of risks from the hypothetical wind farm development. The risk weighting scores from participants were then compared to (correlated with) a set of coded risk attributes based on how interviewees responded to open-ended questions. The topics of these coded risk attributes were derived from the psychometric risk paradigm.

The following subsections explain the study context, sample, interviews, weighting of risk and risk factor calculation methods in more detail.

2.2.1 Study area

New Zealand relies heavily on renewable energy for electricity. As show in Figure 2.1, hydroelectric dams generate a majority of the electricity (57%), followed by geothermal

(16%) then gas plants(12%) (MBIE, 2015). Energy demand continues to increase, as

26 evidenced by consumer energy demand increasing by 4.3% in 2014 (MBIE, 2015). It is possible that expanding electricity production from wind could replace some reliance on fossil fuels, especially if electric cars become more widespread. New Zealand, at latitudes in the “roaring 40s,” has exceptional wind resources (Fortuin et al., 2009), much of which remains untapped. Terrestrial-based turbines in New Zealand generate twice the international average for power generation per turbine (Fortuin et al., 2009). Despite New

Zealand’s abundance of wind, the wind energy sector has been slow to develop (M. Barry and Chapman, 2009).

25,000

20,000

15,000

10,000 (GWh) 5,000

- Annual Electricity Genera/on

Figure 2.1. New Zealand electricity generation by source in 2014 (MBIE 2015). Purple denotes renewable energy sources and gray denotes fossil fuels.

Due to several factors including lawsuits, costs, environmental concerns and public opposition, New Zealand power companies in recent decades have canceled several proposed hydroelectric projects such as the Mokihinui dam (RNZ, 2012a) and wind farms such as Project Hayes in the Lammermoor Range (RNZ, 2012b). Meanwhile, various people and organizations are contesting investments in fossil fuel extraction, as indicated, for example by protesting offshore drilling (NZME, 2015).

In July 2015, the government of New Zealand nonetheless set a national target to “reduce greenhouse gas emissions to 30% below 2005 levels by 2030” (Ministry for the

Environment, 2015). Achieving this goal will requite additional development of low carbon energy. We explore perceptions of concerns and benefits that could be associated with the further expansion of renewable energy infrastructure.

To explore these perceptions, we selected the coastal communities in Golden and Tasman

Bay, New Zealand (see Figure 2.2) in collaboration with Cawthron Institute, an independent New Zealand science organization that conducted a marine ES assessment for this region. Given its relatively shallow water and strong, consistent wind, parts of

Golden Bay, New Zealand are physically well-suited for an offshore wind farm (Fortuin et al., 2009). Suitable locations based on wind strength and water depths are within 20km of Farewell Spit, a 26km long sand bar that is protected as a “Wetland of International

Importance” by New Zealand’s Department of Conservation (Davidson et al., 2011).

The site of the hypothetical farm is physically well suited for the technology, but there are no proposals for wind farms on the site. Despite this, interviewees may have been aware of the controversy leading to the cancelation of a proposed wind farm called

Lammermoor in Central Otago (RNZ, 2012b) and the Makara wind farm near

Wellington, which was built at a smaller scale than originally proposed (O'Neil, 2015).

Figure 2.2. Study Site: Golden Bay, New Zealand. Google Earth image of New Zealand. Public domain. Inset image of Farewell Spit from NASA.

2.2.2 Interview sample

We interviewed people with professions and/or livelihoods linked to the marine environment or energy sector who therefore have a vested stake in marine and energy- related decision-making. Local staff at a research institute (Cawthron Institute) that specializes in marine, coastal and water resources, recommended opinion leaders, business owners, managers and engaged citizens in the region for interviews. We used non-proportional quota sampling (Tashakkori and Teddlie, 2003) to solicit a range of attitudes and opinions with those who have knowledge of marine ecosystems, energy systems, and/or environmental planning. Interviewees worked in the following sectors: fisheries, aquaculture, ecotourism, community planning, environmental consulting, town council, Department of Conservation (government), energy and Maori resource management. Maori are New Zealand’s indigenous people who, based on statute, have

29 environmental interests that must be taken into account by those making decisions about environmental management. A total of 27 people were interviewed, including 18 men and

9 women. A total of 25 were Caucasian and two Maori.

2.2.3 Interview design

The semi-structured interview was designed to identify and weigh the perceived risks and benefits associated with an offshore wind farm. Our methods probed people’s perceptions of the ways in which this hypothetical wind farm might alter the provision of ES.

Interviews began with a warm-up consisting of questions related to occupation and town of residence. Basic information about New Zealand’s current sources of electricity was provided including tables about consumer energy demand by sector. The interviews included the following statement about the context for a hypothetical offshore wind farm:

“electrifying the transportation sector could reduce carbon emissions. This would entail developing additional sources of low carbon electricity.” We then asked questions about perceptions of energy security, attitudes towards existing and proposed renewable electricity sources, and perceptions of risk to energy infrastructure associated with an earthquake, a salient concern in this region. We asked if respondents had heard about offshore wind farms as well as about their concerns and potential benefits of this technology. We also asked if and how people felt attached to Golden Bay. See Appendix

A, B and C for the interview consent form, request letter and protocol, respectively.

The interviewer then showed a three minute animated seascape visualization of an offshore wind farm in Golden Bay created with Google Earth and SketchUp (See

30 Appendix E and https://youtu.be/w_JYLRHi_Bc). Animated visualizations often engage more complex dimensions of perception and aesthetic preference than photographs and text (Sheppard and Cizek, 2009).

Figure 2.3. A still image from the animated seascape visualization of an offshore wind farm in Golden Bay, New Zealand using Google Earth.

After showing the visualization, we provided open-ended opportunities for interviewees to consider potential impacts to ES. We asked: “If you think about the ways in which nature and this place is important to you, what do you think could be lost if this project was developed?” then “What do you think could be gained if it went through?” These results were coded as explained below in section 2.3.2.

2.2.4 Weighting of concerns

Next, interviewees were asked to distribute 20 tokens representing concern across 16 possible topics derived from the literature on public acceptance and rejection of offshore wind farms (Devine-Wright, 2005; Firestone and Kempton, 2007; Gee and Burkhard,

2010; Wolsink, 2010) as well as feedback from local environmental planning

31 practitioners on early drafts of the interview protocol. This list included an “other, specify” category for interviewees to add additional concerns (Table 2.1). We recognize that a few of these topics are interrelated, to the extent that the source of the problem might be one and the same. But all are discrete in relation to the endpoint concern. For example, concerns about property values and tourism are likely related to visual impact.

However, we wanted to know how people assigned weights to and thought about these as parsed topics. Hence, we have a relatively wide range of specific concerns. The list of concerns included multiple impacts to (and/or concerns about) the provision of ES likely to be affected by our proposed wind farm. The ES concerns were not differentiated from the human safety and economic concerns when presented to participants.

Table 2.1. Common concerns associated with offshore wind farms. They were derived from literature and early tests of the interview protocol. Interviewees allocated 20 tokens representing relative level of concern across these topics. The ecosystem service concerns are the dependent variables in the proceeding analysis.

Ecosystem Service Concerns Human Safety & Economic Concerns Potential consequences of risk from wind Potential costs and hazards associated farms to ES provision with a wind farm ______Negative impact on birds Navigational safety issue Negative impact on marine mammals Cost of construction Displacement of commercial fishing Cost of compliance with regulations Negative impact on tourism Cost of maintenance Displacement of recreational fishing Increased cost of electricity Displacement of recreational boating Decreased property values Negative visual impact Insufficient local benefit Negative impact on other species Other (specify) (Specify)

2.2.5 Risk factor scoring using risk perception theory

We investigated the level of concern that people have regarding the potential consequences of an offshore wind farm based on a set of attributes identified in the psychometric risk paradigm. We used the coding scheme in Figure 2.3 to evaluate interview content and literature on offshore wind. An enduring finding in the risk literature is that two fundamental factors drive perceived risk. These are referred to as

“dread risks” and “unknown risks” (P. Slovic, 2010; 1987). Each of these factors generally comprises several qualities, defined in Table 2.2. For example, dread risk is a summative label for whether or not people perceive a risk object as relatively dreaded, controllable, equitable, and/or reversible (see Table 2.2). Instead of using conventional risk research methods of asking people to rate factors such as controllability, we used an open interview design to avoid pre-assigning any logics to how people explained their perceptions of impacts and benefits. We then assigned a risk factor score to each of those

33 concerns. In this sense, we inferred a risk factor score when coding qualitative responses from open-ended questions in the interviews. If attitudes towards a particular risk dimension did not frequently arise in our semi-structured interviews, we relied on academic literature on perceptions of risk related to offshore wind. For each component of the psychometric risk paradigm, we assigned a -1 or +1 as shown in Table 3 and

Appendix D. For example, we assigned +1 to the dread impact to seabirds because people articulated negative emotions and/or dread associated with potential harm to birds. As an example, a man in his late 70s who volunteers for forest and bird conservation efforts said the case study region is a destination for migratory birds. He said, “I would feel dreadful if we suddenly developed [a wind farm resulting in] carnage of those birds....

They do arrive in thousands, tens of thousands."

We scored some categories based on our interpretations of how offshore wind planning processes have unfolded in Northern European countries and the U.S. because the topic did not arise in our interviews about Golden Bay. For example, we scored displacement of commercial fishing as -1, which denotes that it is “controllable” because various stakeholders generally have opportunities to play a role in ocean planning processes and may influence the location and size of a wind farm (Nutters and Pinto da Silva, 2012); people tend to have some control in relation to displacement of fishing.

2.2.6 From scoring risk associated with wind farms to analysis

We then analyzed the assigned risk scores to determine if they correlated with the weights that interviewees assigned to each concern. That is, the composite risk factor

34 score was the explanatory variable and the mean token allocation to the concerns in

Table 2.1 was the dependent variable. This partially tests the extent to which perceived intensity of risks to ES provision is predictable based on attributes of the psychometric risk paradigm.

Table 2.2. Psychometric Risk Characteristics (P. Slovic, 1987). The left component of each pair reduces risk perception while the right increases it. Subcomponents in gray italics varied little across the ES concerns reported by interviewees so were not included in the analysis. See Appendix D for full explanation.

Factor 1 Dread Risks Factor 2 Unknown Risks Controllable Uncontrollable Risks known to science Risks unknown to science Not dread Dread Observable Not observable Consequences not fatal Consequences fatal Known to those exposed Unknown to those exposed Equitable Not equitable Eﬀect immediate Eﬀect delayed Easily reduced Not easily reduced Old risk New risk Not globally catastrophic Globally catastrophic Low risk to future High risk to future genera7ons genera7ons Risk decreasing over 7me Risk increasing over 7me Voluntary exposure Involuntary exposure

35 Table 2.3. Explanation of composite risk factor scoring. Some risk characteristics are associated with consequences to ES from a wind farm. The psychometric risk paradigm inspired our risk characteristics (components of Factor 1 and 2). Our scores in gray are based on data from our interviews and publications from the social and ecological sciences on offshore wind farms. We removed several factor 1 and 2 characteristics (e.g., globally catastrophic, risk to future generations) because that appeared to vary little across the ES concerns. WF is wind farm. A score of 1 means this increases perceived risk while a -1 means it diminishes perceived risk according to the risk perception literature. See Appendix D for further explanation of omitted risk characteristics. Factor 1 Dread

Can the person who suffers Does potential consequence Is a particular Risk factor negative consequences control the Can precautions be easily taken to reduce the negative impact? evoke a feeling of dread? consequence fatal? severity of the consequences?

Diminishes risk Consequences not fatal (- Controllable (-) Not dread (-) Easily reduced (-) perception (-) ) Car: driver can drive cautiously to Example reduce severity of potential Bicycle, car Medical x-ray Medical x-ray: wear a lead apron, bicycle: wear a helmet accident Increases risk Uncontrollable (+) Dread (+) Consequences Fatal (+) Not easily reduced (+) perception (+) Airplane: passengers relinquish Terrorism, shark attack, nuclear Example control to pilot, passengers do not Nuclear meltdown Ocean acidification meltdown control severity of accident

Displacement of recreational -1 -1 -1 -1 fishing Stakeholders generally have Area displaced tends to be opportunities to influence location relatively small in comparison As long as area of wind farm is not prime or irreplaceable fishing and size of wind farm; they tend to the much larger extent of Not fatal grounds, impact can be reduced by moving fishing effort to have some control in relation to fishing grounds, this tends not elsewhere displacement and consequently to be not a dreaded concern impact on fishing Displacement of commercial -1 -1 -1 -1 fishing Area displaced is small relative Same as above in relation to Impact easily reduced by moving commercial fishing effort to size of bay, this is not a Not fatal commercial fishing elsewhere dreaded concern Displacement of recreational -1 -1 -1 -1 boating No expressions of dread found Same as above in relation to in literature in relation to Not fatal Impact easily reduced by recreational boating elsewhere impact on fishing displacement of recreational boating Negative impact -1 -1 -1 1 on tourism

Results are inconclusive regarding if wind farms No expressions of "dread" per negatively impact tourism. It is a se found in literature in relation common concern, but tour to negative impact on tourism Not easily reduced: tourism operations would likely need to operators control what they nor in interviews. People are Not fatal change their operations that currently focus on wildness of land advertise and show so they could concerned, but we did not find and seascape capitalize on the green tech aspect documentation of widespread of farm. Many tourists may want anxiety or fear (aka dread). tours of the farm (Lilley, 2010).

Negative visual 1 -1 -1 1 impact Potential Ecosystem Service Consequence Ecosystem Service Potential Dread or fear does not The negative affective reaction to characterize most people's Placing the turbines further offshore to reduce visual impact is not visual impact is subjective. We attitudes to a WF. Many dislike Not fatal feasible with existing technology given water depths at distances interpret it as uncontrollable. and don't want it but it's not a at which farm would not be visible from land source of dread. Impact on 1 1 1 1 seabirds People strongly value region's Extensive studies on bird migrations have been conducted to People tend not to control bird high density of nesting sea Some bird mortalities are inform siting of WFs. Once constructed, few options currently behavior. Perception of high birds. There is widespread fear associated with wind exist to reduce risk of bird collisions with commercial scale likelihood of collisions that development could harm turbine collisions modern turbines bird populations. Impact on marine 1 1 1 1 mammals People dread potential harm to Perception of fatal whales as evidenced by strong collisions (although none Can not control marine mammal affective response in interviews have been documented in behavior with regards to wind and to whale strandings and WF studies); perception Interviewees do not know of technologies to safely keep whales turbines, collision is a common deployment of volunteer time that electromagnetic fields away from turbines concern and resources to reduce from underwater cables fatalities of common whale could effect whale strandings in bay strandings

2.2.7 Weighting of benefits

We also asked participants to weight the benefits associated with a potential offshore wind farm by allocating 20 tokens across 16 potential benefits, including an “other, specify” category.

These benefits, which accrue at different scales (local, regional, national, global), were based on renewable energy literature (Dincer, 2000; Snyder and Kaiser, 2009).

2.3 Results

Our results suggest that particular risk dimensions from psychometric risk theory are positively correlated with the mean level of concern to risk items that our study participants assigned.

Results include narrative expressions of concerns as well as benefits and an exploratory quantitative analysis of our data on wind farm concerns.

2.3.1 Concerns

2.3.1.1 Narrative expressions of concern

Interviewees expressed “place-protective” concern about development of any kind in this area, similar to concerns expressed in Devine-Wright (2009). For example, one explained that “our marine environment is not a built environment and you are extending the built environment… beyond the land. Personally in an ideal world I wouldn’t want to see the built environment extend around the coasts...extending the built environment into the marine area. That is impacting on the… the wilderness, intrinsic values… And it is Farewell Spit.”

37 Interviewees expected the magnitude of impact to sea birds and visual impact and to be large.

One interviewee said, “I think the bird kill from those wind farms is massive isn’t it?”

Interviewees had contrasting affective reactions to the visual impact. One said, “Something like that would completely alter the view.” An interviewee involved in local tourism and government complained that the visualization “makes me feel sick…. Something like that would completely alter the view…half the population [of Golden Bay] would think it would be a really good idea and the other half would think it is a bloody disaster.”

A local government consultant said wind turbines “look stunning…they are quite a striking feature… It’s easy to look at a wind farm in someone else’s back yard and say it looks stunning and that it is a great place for it but if there was a proposal for a wind farm out there [near

Farewell Spit], no I don’t think I would support that. I would rather see one somewhere up on the hills on the back here.”

A mid 60-year old female policy planner said “the negative impact is more than visual and ecological. There is a component to landscape that is to do with a sense of place, a sense of associations and meanings. So it’s the cultural, it’s how it is interpreted through art and aesthetics. So it’s more than visual. Visual you might just be looking at it purely in terms of aesthetics but it is the meaning that people hold.”

38 2.3.1.2 Weights assigned to concerns

The data on concerns address the first research question. As shown in Figure 2.4, interviewees assigned the highest level of concern to the potential impact of the wind farm on birds, followed by negative visual impact and impact on marine mammals.

Impact on birds* NegaGve visual impact* Impact on marine mammals* Cost of construcGon Insufficient local benefit Displacement of commercial fishing* NegaGve impact on tourism* NavigaGonal safety issue Cost of compliance with regulaGons Cost of maintenance Displacement of recreaGonal fishing* Increased cost of electricity Other:Specify Displacement of recreaGonal boaGng* NegaGve impact on other species* Decreased property values 0 1 2 3 4 5 Mean of Rela/ve Concern

Figure 2.4. Relative weighting of offshore wind farm concerns with standard error. Participants distributed 20 tokens representing the weight of their concern across 16 topics. The concerns denoted with a “*” can be understood as potential consequences of the risk from wind farms to the provision of ES. The relatively low assignment of concern to the “other” category indicates that our specified categories captured the vast majority of what people worry about in relation to this hypothetical context.

We ran a correlation to explore the second research question on the extent to which psychometric risk dimensions and the associated affect heuristic predict how study participants weight various

ES concerns. The psychometric risk paradigm predicts that some risks are perceived as higher than others based on a relatively small subset of risk dimensions (e.g., perceived dread, controllability, fatality of consequences, reducibility of consequences). We used these

39 dimensions of perceived risk from the literature as discrete predictor ‘psychological’ variables.

The dependent variable is the weight assigned to relative concern for specific ecosystem services

(e.g., concern about tourism, visual impact, sea birds). As show in Figure 2.5, the composite risk factor score positively correlates with the mean level of concern that interviewees expressed when they assigned tokens to various ES impacts (R2 = 0.67).

5 4.5 4 Seabirds

3.5 R² = 0.67007 3 Visual Impact 2.5 Commercial Fishing 2

1.5 Marine Mammals Tourism Rec Fishing 1 0.5

Mean Level of Concern for Ecosystem Services Rec BoaGng 0 -5 -4 -3 -2 -1 0 1 2 3 4 5 Composite Risk Factor Score: Perceived Dread, Controlability, Consequences, Reducability

Figure 2.5. Average level of concern according to stakeholders for risks to ES plotted against psychological dimensions of each risk with standard error bars. The level of stakeholder concern was quantified as the mean number of tokens representing amount of concern that interviewees assigned to potential consequences of an offshore wind farm to the provision of various ES. The composite risk factor score expresses the nature of the risk, not its magnitude, via a set of risk dimensions. These risk dimensions were determined based on published literature and qualitative responses during the interviews on if each ecosystem service impact is generally perceived as uncontrollable, dreaded, has fatal consequences and is not easily reduced. See Table 2.3 for more explanation on the scoring.

The distribution of tokens allocated to different concerns can be interpreted as allocations proportional to the magnitude of expected impacts (see Discussion). Given the small footprint of the wind farm relative to the bay, the expected magnitude of the impacts are small to ES that may be displaced, such as fishing and boating. People expressed considerable uncertainty about

40 the impact of the farm to marine mammals. A study participant who was a technical adviser for marine resource management said “when it comes to marine mammals, I’ve got no idea what that [wind farm] means for them… Will those towers be perceived as a threat? Will they see them as a curiosity? Will they attract more [whales]…Or it could be a negative…. I’ve got no idea.”

2.3.2 Benefits

2.3.2.1 Narrative expressions of benefits and trade-offs

Some interviewees emphasized positive rather than negative aspects of visual impact. A male environmental planner in his late 40s said, “I don’t see them as threatening, I just see them as an opportunity… that you need to work through and figure out how the community is going to react to it, how much they appreciate and understand local [electricity] supply requirements, whether they are willing to accept their own footprint in their own backyard…. I like people to see where it’s [electricity is] coming from and having the effect localized.” He assumed that people would be more responsible energy consumers if they lived proximate to their sources of electricity and saw the environmental impacts of their personal electricity consumption regularly.

In stark contrast to the more common negative perception of visual impact, a female environmental planner in her early 50s said offshore wind turbines “are amazing. I don't find them offensive at all. I think they are quite beautiful… they're almost like a sculpture… in the right context they're quite neat.”

Two interviewees responded positively because an offshore wind farm aligned with their work to

41 create more marine reserves. A mid-40 year old tourism company owner who has been actively engaged in marine conservation issues referred to the potential displacement of commercial fishing by a wind farm in a positive way. He said marine reserves “work so well and there’s so little of them. I know everyone’s got to eat and people have got to make money but I think it’d be pretty awesome to see a cluster of windmills there to protect that environment under it.”

A female in her late 70s who was an active volunteer for an environmental advocacy group expressed a willingness to accept a view with anthropogenic structures as long as it contributed to reducing reliance on fossil fuels: “Your example of Farewell Spit, that’s something that is very precious to me and I would be prepared for there to be a windmill there if it is as you say one of the best places in NZ for [offshore] wind. So it’s iconic to me but I could still accept a windmill for the sake of not having to use petrol.”

2.3.2.2 Weights assigned to benefits

We addressed the third research question using benefit-weighting data to determine the relative importance assigned to benefits associated with an offshore wind farm. The most heavily weighted benefit associated with this hypothetical wind farm was increased regional self- sufficiency, followed by increased diversity of New Zealand’s energy portfolio, then the contribution to New Zealand’s energy independence. See Figure 2.6.

42 Increased regional energy self-suﬃciency

Increase diversity of New Zealand’s energy poraolio

Contribute to New Zealand energy independence ReducGon in carbon emissions associated with electricity generaGon Source of new jobs in Golden Bay

A source of local pride in energy innovaGon Benefit to recreaGonal fishing because fish will aggregate near structures Electricity without natural resource depleGon

Increased local control of energy producGon

Taps into abundant local resource (wind)

Electricity without impact on air quality Increase marine species abundance from arGﬁcial reef eﬀect This could start a new industry in New Zealand

PosiGve visual impact

PosiGve impact on tourism

Other: specify ______

0 1 2 3 4 5 Mean of Rela/ve Beneﬁts

Figure 2.6. Perception of relative value of benefits from an offshore wind farm. The mean number of tokens assigned to each benefit is presented with the standard error.

2.4 Discussion

Participants assigned higher levels of concern to affectively loaded topics, including visual impacts and impacts to iconic wildlife, than the topics typically included in cost benefit analyses

(see Figure 2.4). We found a large positive correlation between the elements of the psychometric risk paradigm, which we incorporated as composite risk factor scores, and how interviewees assigned tokens representing relative concern for different risks to ES (see Figure 2.5, R2 = 0.67).

43 Based on this correlation, our analysis supports an expansion of the predictive power of the psychometric risk paradigm beyond its original focus on risks to human health and safety. Our findings suggest attributes from the psychometric risk paradigm, specifically notions of control, dread, associated fatalities of animals and the reducibility of a risk, can help predict relative levels of concern associated with the consequences of renewable energy, in this case an offshore wind farm. Local residents tended to express greater relative concern regarding potential losses of ES that tend to be uncontrollable, dreaded, associated with animal fatalities and irreducible.

This means, for example, that high concern for birds could have been anticipated based on the dimensions of the psychometric risk paradigm. It would thus appear that theories of risk that have been powerful in explaining variation in perceptions of personal harm also apply to indirect risks experienced via ES (e.g., bird strikes).

Our results have implications for wind farm developers, wind farm regulators and, more generally, for people conducting assessments relevant to a proposed change in ES provision. We recommend that developers, regulators and ES assessors pay attention to concerns voiced by stakeholders characterized by attributes of the psychometric risk paradigm about potential changes to the delivery of ESs. In our case study, potential impacts of offshore wind farms on birds and marine mammals and negative visual impacts emerged as the top concerns. These top concerns are most closely linked to attributes long associated with high levels of perceived risk.

Addressing such concerns could entail acquiring additional scientific information, e.g., conducting environmental impact analyses of wind farms on birds and marine mammals (a requirement in most developed countries), a trade-off analysis for reducing visual impact (i.e.,

44 develop scenarios for siting farms at different distances from shore while recognizing the higher costs further from shore), and developing mitigation strategies for these potential impacts (e.g., funding bird habitat restoration elsewhere). Conducting such analyses, however, is not sufficient for addressing local concerns. The scientific results and mitigation plans ought to be communicated in publically accessible ways (see Klain et al., 2015) and incorporated into a deliberative decision process. Such a process would include negotiations about relevant facts and values, particularly those strongly associated with attributes of the psychometric risk paradigm.

Our results also demonstrate that, among our sample of people with natural resource and energy related livelihoods, environmental concerns tend to be weighted more heavily than economic costs as shown in Figure 2.4. For example, people on average expressed a higher level of concern about impacts on birds and marine mammals than the costs of construction, compliance with regulations, and maintenance as well as the increased cost of electricity. We acknowledge that this project was hypothetical with no real costs to be borne by participants. The results, however, do align with findings from Ansolabehere and Konishky (2014), who conducted surveys of U.S. citizens demonstrating that people want clean and cheap energy, but the foremost concerns driving energy preferences are minimizing environmental harms, then economic costs.

We recognize limitations to our study. We do not compare our results to objective metrics associated with risk to ES. Instead, we used our understanding of pertinent literatures to derive composite risk factor scores. We used our coded risk factor statements and scores (independent variable) to understand the logics behind the weighting of concerns associated with a potential wind farm (dependent variable). One limitation with our method is that it is possible that

45 variation in the perceived absolute magnitude of the risk contributed to the observed relationship between the coded risk factor scores and relative levels of concern.

Future research could address similar questions using survey methods rather than interviews to obtain a larger and representative sample. Nonetheless, the qualitative data from these interviews illustrates the diversity of values and risk perceptions associated with this novel technology.

Our case study demonstrates how offshore wind can be an ambiguous risk, which means there are various legitimate perspectives on the extent to which the technology may result in adverse impacts, largely due to scientific uncertainties. Also, ambiguity arises when there is no consensus as to whether potential impacts are acceptable, tolerable or intolerable (Renn et al., 2011).

Another contributor to ambiguity is that people respond to risks based on their particular risk- related images and constructs (Keeney, 2004). If a wind farm was proposed in our case study area, we would expect numerous legitimate interpretations of results from any formal risk and/or

ES assessment given the ambiguities associated with risks related to offshore wind.

Participants in our research, as demonstrated in the narrative results, voiced concern for biodiversity in relation to the hypothetical wind farm. This concern points to a need for future research and potentially scenario-based ecosystem service assessments. This work could investigate the extent to which designing renewable energy infrastructure that also provides natural habitat (e.g. an offshore wind farm built with excellent artificial reef habitat) can elicit positive affective responses that could neutralize negative affective responses to this technology.

46 2.5 Conclusion

The affectively-loaded language and risk ratings evident across stakeholders’ evaluations of the benefits and concerns about an offshore wind farm have implications for energy transitions.

Traditional risk assessments quantified biophysical and economic risks, which tended to be reduced to estimates of probability and severity. They overlooked critical psychological dimensions of risk (NRC, 1996). Affectively charged dimensions of risk can profoundly impact the uptake of new technology, how ES valuations are interpreted and consequently how society deals with climate change.

Our results have implications for communications used to introduce proposed wind farms, particularly those directed to communities near a proposed development site. We recommend that such communications anticipate and be sensitive to perceptions of control, dread, associated fatalities of animals and the reducibility of risks associated with a proposed development.

Identifying and disseminating mitigation measures for concerns associated with these qualities could help garner greater public support for renewable energy developments.

Based on our results and the social science literature on risk, we recommend that energy infrastructure proponents invest considerable effort into interdisciplinary and deliberative risk estimations that support mutual learning among diverse constituents, which is necessary for managing uncertain, complex and/or ambiguous risks. Deliberative, participatory processes can account for a diversity of causal beliefs embedded in different worldviews in relation to risks. In the words of Renn (2011, p. 240), “what is safe enough implies a moral judgment about

47 acceptability of risk and the tolerable burden that risk producers can impose on others.” Our research extends these “others” to include non-human species.

Our research also has implications for initiating broader discussions on re-imagining our energy systems to transition towards reliance on low carbon renewable energy rather than fossil fuels.

We recommend that planners and proponents of renewable energy technology pay particular attention to the ways in which the consequences of our current energy systems and proposed renewable energy developments can be interpreted, amplified or played down by the public and stakeholders in relation to the attributes of the psychometric risk paradigm.

48 Chapter 3: Rethinking renewable energy: high willingness to pay for ecologically regenerative offshore wind farms

Sarah C. Klain, Terre Satterfield, Kai M.A. Chan

3.1 Introduction

The ongoing quest to define and secure sustainable energy is one of humanity’s most pressing challenges, particularly in the context of climate change (Yergin, 2011). The United Nations set ambitious sustainable development goals, including universal access to affordable, reliable, sustainable and modern energy (UN, 2015) while the World Bank and International Energy

Agency’s Sustainable Energy For All initiative calls for the doubling of renewable energy in the global energy mix and tracking this energy transition (Angelou et al., 2013). One approach for mitigating climate change involves rapidly scaling up low-carbon energy production to replace energy from fossil fuels. Various pathways have been proposed to transition away from fossil fuels and towards renewable energy, with offshore wind playing a substantial role in proposed pathways for several countries with coastlines (Foxon et al., 2010; Green and Vasilakos, 2011;

Jacobson and Delucchi, 2011). Along the US Eastern seaboard, Kempton (2005) argues that offshore wind is the only spatially proximate utility-scale renewable energy source that could displace significant carbon emissions in the near term.

And yet, negotiating what constitutes both clean (i.e., no greenhouse gas emissions) and locally desirable energy systems is an ongoing debate at local and regional scales with global ramifications (Devine-Wright et al., 2011; Roberts et al., 2013), despite the fact that scientific

49 consensus on the need to decrease greenhouse gas emissions has coalesced. Also, environmental and human safety risks associated with large scale offshore wind farms (OWFs) over their life cycles are relatively benign as compared to risks associated with other energy sources, including fossil fuels and nuclear (Ram, 2011). For example, OWFs, in contrast to coal plants or nuclear reactors, do not pose any catastrophic risks that could result in human deaths or property damage in excess of $1 million (Ram, 2011).

Despite this relatively low level of risk, developing offshore wind farms has been controversial for various economic, social and environmental reasons. In particular, wind farm debates have focused on their relatively high levelized costs compared to fossil fuels; dissatisfaction with who owns and operates these utilities; visual/aesthetic impact of the farms; and impacts on species and habitat (Firestone and Kempton, 2007; Firestone et al., 2012; Pasqualetti, 2011; Wiersma and Devine-Wright, 2014). In light of these concerns, our research estimates how specific wind farm features augment or erode public support for developing this technology where each feature is also linked to a cost, measured as willingness to pay.

We see the potential for regenerative design (Lyle, 1996) to reduce negative perceptions of some technologies. Design in this context refers to the confluence of society and technology in the conception and shaping of systems. Regenerative design refers to planning and implementing systems that evolve from their initial forms and renew a site, thereby shifting it to a more ecologically desirable condition via human intervention. Lyle (1996) characterizes regenerative systems with the following attributes:

50 • System operations are integrated with natural and social processes • Minimal reliance on fossil fuels and synthetic chemicals • Minimal use of non-renewable materials • Sustainable use of renewable resources for operation • Associated waste products are re-assimilated without environmental harm

The materials to build OWFs, particularly the foundations and submarine cables, are energy intensive and many are non-renewable. Lifecycle analyses, however, demonstrate that, the

“payback” period of properly-sited large-scale wind turbines as related to energy and greenhouse gas emissions embedded in the materials relative to the electricity they generate, is less than one year (Wagner et al., 2011). The global warming potential per unit of electrical energy generated by wind farms is lower than solar photovoltaic, biomass and fossil fuel energy sources (Weisser,

2007).

In the context of wind farms, then, regenerative design might address various ecological risks and uncertainties, including collision risk and diversion of migration routes for seabirds

(Kuvlesky et al., 2007) and potentially bats (a problem for terrestrial farms, but little documentation exists for offshore sites) (Arnett et al., 2008). In the construction phase, acoustic disturbance from pile driving likely has a high impact on marine mammals, fish and benthos

(species inhabiting the seafloor). Moderate to high uncertainty is associated with acoustic impacts to wildlife during the OWF’s operational phase (Bergström et al., 2014).

Electromagnetic fields are anticipated to have a relatively low impact on marine species, but this impact remains uncertain (Bergström et al., 2014; Gill, 2005). Uncertainty about ecological impacts remain given the relative novelty of industrial scale OWFs and how few studies have

51 assessed the cumulative impacts and long-term food web effects associated with them

(Bergström et al., 2014; Goodale and Milman, 2014).

Ideally, offshore wind farms could go beyond mitigating negative impacts to instead benefit or enhance marine habitats. Human activities have strongly affected approximately 41% of the ocean on a global scale (Halpern et al., 2008). Dredging, mining and some fishing practices, such as bottom trawling, have reduced benthic structural diversity, which diminishes habitat complexity, thus altering species composition and diversity (Auster and Langton, 1999; Watling and Norse, 1998). Reversing these trajectories has the potential to increase localized biodiversity while ecologically benefiting the surrounding marine environment through appropriate design, better siting and management and artificial reefs (Baine, 2001; Bohnsack and Sutherland, 1985).

OWF turbine foundations could act as artificial reefs and fish aggregation devices, both of which have contributed to restoring degraded marine ecosystems (Boehlert and Gill, 2010; Inger et al.,

2009). Further, an OWF may become a de facto marine reserve with associated conservation benefits (Bergström et al., 2014; Pelc and Fujita, 2002).

For these reasons, this study examines the degree of public support for wind farm design expressed as support for particular regenerative effects, and asks whether positive attributes can outweigh negative impacts (heretofore the primary focus of research). We assess this support by quantifying the attributes linked to stakeholder preference for one potential wind farm over another. OWF biodiversity benefits may largely accrue underwater where they are not readily visible. In contrast, the highly visible turbines may impact some bird species and degrade the perceived aesthetic quality of a seascape. We operationalize debates about perceptions of wind

52 farms as diminishing the aesthetic quality of a land or seascape, typically referred to as a negative externality (Devine-Wright, 2005; Ladenburg and Dubgaard, 2007; Warren and

McFadyen, 2010), while noting the rising cost per unit of energy generated as wind farms are sited further from shore (Snyder and Kaiser, 2009). We further test willingness to pay for more distant and consequently less visible OWFs (Krueger et al., 2011; Ladenburg and Dubgaard,

2007; Westerberg et al., 2013), and we examine the effect of increased cost as linked to increased biodiversity benefits.

In addition to distance from shore, ownership can also exert a significant impact on wind farm preferences. Using choice experiment methods, Ek and Persson (2014) found that Swedish residents prefer cooperatively or municipally owned wind farms over private and state-owned farms. OWF ownership preferences have yet to be assessed in our study area of coastal New

England (Figure 3.1), where cooperative ownerships models are common in some sectors of the economy, e.g., lobster and fisheries cooperatives (Acheson, 2003). An energy cooperative exists in this region (Vineyard Power) and is partnering with an developer to potentially become a part owner of an OWF (Nevin, 2010). We thus operationalized ownership model variables as well

(see also Ek and Persson, 2014).

In order to prioritize wind farm characteristics that might make its development more socially acceptable, we quantify preferences for wind farm attributes using a choice experiment. We used an online panel of residents from coastal New England states where, as of 2016, North

America’s first OWF is under construction. We estimate how much public support there could be for ecologically regenerative effects, and whether such positive attributes can outweigh negative effects. This understanding yields estimates of WTP for an OWF that provides marine

53 biodiversity benefits via habitat provision, as well as quantified public preferences regarding visual impacts and ownership type. This is the first study to assess these features concurrently and in this geographic area.

3.2 Methods

Our methods included three major components. First, we recruited respondents using Amazon’s

Mechanical-Turk platform, restricted to residents of our study region. Second, we presented respondents with a choice experiment offering options of wind farms vs. a default of fossil fuel electricity generation, complete with visuals. Third, we used standard econometric analysis to infer from respondent choices the relative preference for different OWF attributes (e.g., near vs. far from shore; biodiversity losses vs. gains, private vs. cooperative ownership), and the WTP for levels of those attributes. We explain each of these components in greater detail below.

3.2.1 Study location

New England coastal states have strong and consistent wind resources offshore (see Figure 3.1).

An energy transition towards greater reliance on renewables would likely include hundreds of

OWFs off the coasts of these states (Jacobson et al., 2015a; 2015b). Large-scale OWFs have been developed in Northern Europe, but only one small farm near Block Island has been built in

North America as of 2016. We chose to assess public preferences related to OWFs based on a survey of coastal New England residents because of this region’s high wind resource potential and the fact that several farms are currently under consideration near the coasts of these states.

54 Quebec

New Brunswick

Maine

Vermont

New Wind resource York New Hampshire poten4al Poor

Fair Massachuse1s N Good Excellent Rhode Connec4cut Island Outstanding

Figure 3.1. Wind resource potential for states in study. Wind data from NREL (2015).

We tested a pilot of the choice experiment with 20 individuals to ensure that the survey was clear. We made minor adjustments to clarify the wording of the survey (see Appendix F, G and

H for the survey consent form, the M-Turk request description and the survey respectively).

3.2.2 Sample characteristics

We recruited respondents using Amazon’s Mechanical Turk (M-Turk) system, which has become a common respondent recruitment method for experimental research (Goodman et al.,

55 2012; Paolacci et al., 2010) with data outputs that are as reliable as those acquired via traditional recruitment methods (Buhrmester et al., 2011). We attempted to minimize bias in our sample by describing it on M-Turk’s HIT list (Human Intelligence Tasks) in very general terms (as a survey about preferences based on different text and image-based descriptions, without using any language related to renewable energy). The sample was limited to M-Turk workers who have mailing addresses in coastal New England states (Connecticut, Maine, Massachusetts, New

Hampshire or Rhode Island), where several proposals for OWFs are more advanced than elsewhere in North America.

Respondents meeting the location requirement were provided with a link to our survey hosted on the Qualitrics survey platform. We used the Qualitrics software to randomly assign survey-takers to one of four blocks of choice experiment questions. We collected self-reported demographic data from the sample so we could compare it with census data to determine the extent to which this sample is representative of the population of these states. When they completed the survey, respondents were given a completion code to submit in the M-Turk system for payment.

Respondents were paid $1 to take the 10-15 minute survey.

A total of 412 respondents completed the survey. We excluded data from 12 respondents who failed two questions we inserted to test if the survey takers were paying attention (see Appendix

I, question 6 and 47). This type of screening based on attention-check questions is recommended when relying on ‘Mechanical Turk’ workers (Goodman et al., 2012).

56 Our respondent pool is typical of M-Turk workers as described in Paolacci (2010). Our sample had higher self-reported level of education than the general population, was younger (32 years vs

40 as the mean age in these states), more females (59%) than males (41%), and self-reported household income lower than the states’ average (Table 3). The white/non-white racial breakdown of our sample (82.5% white) corresponded closely to census data (82.2% white).

Table 3.1. Survey respondents demographic characteristics compared to census data.

Socioeconomic Percentage or Percentage or Mean Characteristics Description N Mean of Sample from 2014 Census* Bachelor degree or Education higher 400 66.3% 37.9% Age Years old 400 32 40 Female Gender 400 59.0% 51.3% Annual household Income income before taxes 400 ~$53,000* $66,200 State CT 400 18.5% 25.6% ME 400 12.3% 9.5% MA 400 45.5% 48.0% NH 400 9.5% 9.4% RI 400 9.8%** 7.5% White Caucasian race 400 82.5% 82.2% We used 2014 Census data from coastal New England states including Maine, Massachusetts, New Hampshire, Connecticut, Rhode Island. State census data was summed then weighted by the state’s population size.

* This is approximate because survey respondents selected an income category rather than reported a specific amount. For example, category 5 corresponds to $35k to $49k while category 6 is $50k- 74K. The mean income was 5.4, which we interpret as 40% of the value between the middle of category 5 and 6 at ~$53,000. ** All respondents have a mailing address in a coastal New England state (a requirement for eligibility to take this survey), but 4.4% of the sample did not self-report a zip code in one of these states.

3.2.3 Choice experiment design

In a choice experiment, respondents are presented with options that include various attributes and they are asked to select their preferred option. The attribute levels are varied based on experiment design rules so that researchers can build models for choices based on the attributes of the option that respondents selected. 57

We developed and implemented an online survey hosted on the survey platform Qualtrics. Each survey included a university-required research ethics consent form, introductory material on

OWFs, the choice experiment component, then demographic questions and finally questions about environmental values. The minimum number of attribute level combinations required to estimate orthogonal main effects for four levels each with four attributes was 32, which we divided into four blocks of 8 choice sets. We used a fractional factorial design (Louviere et al.,

2000) for our choice experiment component in order to keep the survey short and reduce cognitive burden for the respondents. We used the choice experiment design tool in the software package JMP to generate our fractional factorial design.

We used four OWF attributes in the choice experiment: effect on marine life, type of ownership, distance from shore, and addition to monthly electricity utility bill. Each attribute had four levels

(see Table 3.2 and Appendix I). The increases in species diversity and abundance are based on literature on artificial reefs, wind farms and environmental impacts, which document high levels of variability across sites and species (IUCN, 2010; Reubens et al., 2013a; 2013b). Although

60% decrease and 60% increase to diversity and abundance are more extreme than most anticipated assessments of impact, we contend that such changes are possible, particularly if the base levels of diversity and abundance are low. Our payment vehicle was a monthly addition to the electrical utility bill. The levels of the bill were based on Krueger (2007) and Krueger et al.

(2011), who recommended a fee over the lifetime of a project. This study used a range of utility fees up to $30 a month for three years, but, based on their model outputs, found that many

58 residents were WTP more than this. We used a monthly fee over the lifetime of the project, which we stated as ~25 years (see Appendix I Choice experiment survey).

Table 3.2. Description of attributes and levels used in the choice experiment.

Attribute Description Levels Biodiversity Percent change in marine • 60% decline species diversity and • 30% decline* abundance • 30% increase • 60% increase Ownership Owner of wind farm • State ownership type • Municipal ownership • Private ownership* • Cooperative ownership Distance Distance of wind farm • 1 mile, highly prominent* from nearest shore • 4 miles, prominent • 8 miles, somewhat visible • > 10 miles from shore, barely visible Bill Monthly addition to • $1 electricity utility bill to • $5 fund wind farm • $10 development • $20 *Denotes base case levels

Each survey respondent was asked to assume that his/her state has committed to increasing electricity generation by 10%. Then he/she was presented with 8 choice sets. Each choice set had three options. Similar to Kruger (2011), option A or B were OWFs with different attributes.

Option C, the “opt-out” choice, was for constructing a fossil fuel plant (see Figure 3.2 for a choice set example). We created visual representations of changes to marine life using vector images from the IAN image library (IAN, 2015) of species common in the Gulf of Maine. We used Google Earth, Sketch Up and OWF models (reaching a virtual height of 70m above sea level) from 3D Warehouse to create OWF visualizations at different distances from shore (see example in Figure 3.2). Visualizations, including photo simulations, are a common feature in wind farm preference surveys (Bishop and Miller, 2007; Ek, 2002; Krueger, 2007; Wolk, 2008).

59 Op#on A Op#on B Op#on C Wind Farm Wind farm Coal or Gas Plant No Wind Farm Eﬀect on • Small loss • Large gain • More coal or natural gas marine life • 30% decline in diversity and • 60% increase in diversity and used abundance abundance • No direct impact on marine • Turbine structures provide • Turbine structures provide ecosystems

poor habitat for underwater excellent habitat for • Associated CO2 emissions plants and animals, e.g., an#- underwater plants and contribute to ocean fouling paint used on tower animals acidiﬁca#on

Wind farm Coopera#ve Private Ownership not speciﬁed Ownership

Visibility Prominent Barely visible Built on land from shore 4 miles from shore ≥10 miles from shore

Addi#on to $5 $20 $0 monthly electricity u#lity bill

Figure 3.2. Example of choice scenario. Images made with graphics from IAN image library (IAN, 2015)

60 3.2.4 Econometric analysis of choice experiment data

The Random utility model (RUM) typically underpins choice experiment data analysis

(McFadden, 2001; Train, 2009). This approach assumes that individuals maximize their utility

(satisfaction) when making discrete choices from a set of alternatives for goods and services. The attributes of a chosen option are assumed to generate individual utility. A RUM relates observed or stated choices to this individual utility. The respondent n obtains utility U, which depends on their choice, from an alternative i out of options j such that 1 < i < j in choice task t. The indirect utility function of respondent n is denoted as Unit:

Unit = β’nXnit + εnit

Respondent characteristics (e.g., demographic variables) and observable attribute levels of option j are represented by Xnit. The coefficient vector of these attributes is βn, which may be random or non-random variables. An unobservable random error term is εnit. As described in Börger et al.

(2015, p. 129) the probability Pnit that respondent n chooses alternative i over all other alternatives in choice task t is:

Marginal WTP or the implicit price for an attribute as compared to a particular baseline can be calculated as the ratio between an attribute level’s coefficient and the payment coefficient (e.g., distance of 5 miles from shore coefficient divided by the bill coefficient). This can be interpreted as willingness to make a trade-off between each wind farm attribute (e.g., 5 miles from shore)

61 and a price attribute (e.g., utility bill) as a change from the base level of an attribute (e.g., 1 mile from shore).

Similar to Krueger (2011), we include an “opt-out” choice in the utility function (a fossil fuel plant rather than a wind farm with no additional bill). The attribute levels of the opt-out choice are fixed (e.g., no ownership type specified, no impact on marine biodiversity, no visual impact, no additional cost to utility bill). The logit models incorporate selections of the “opt-out” choice when coefficients and WTP amounts are estimated (e.g., if more respondents choose Option C of a fossil fuel plant, WTP for wind farm attributes decreases).

We used conditional and mixed (also known as random parameter) logit models to infer how respondents value certain wind farm attributes relative to other attributes. The conditional logit model assumes that preferences are constant across respondents. It also assumes that εnit has a type 1 (Gumbel) extreme value distribution. The mixed logit model allows for “random taste variation, unrestricted substitution patterns, and correlation in unobserved factors over time”

(Train, 2009, p. 134). A mixed logit model consists of fixed as well as random effects. Taste parameters—respondents’ personal preferences embedded in the utility components denoted as

βn—vary randomly across the sample population in mixed logit models.

We considered ownership type, distance from shore and impact on marine species abundance and diversity as categorical attributes. Following Louviere et al. (2000), the categorical attributes were used as effects-coded variables. A categorical variable that has n levels is replaced with n -

1 effects-coded variables. We refer to the omitted level as the base case. The significance of

62 coefficients of other levels are relative to the base case levels, which are noted in Table 3.2. We chose our base case attribute as the shortest distance from shore (1 mile) because people tend to derive greater utility from less visible wind farms. We also selected small loss to diversity and abundance as the base case and private ownership, which was arbitrary (see Table 3.2). “Bill”, the payment mechanism, was used as a continuous variable in the models.

3.3 Results

Our choice experiment results (Table 3.3) show that the strongest preference for the OWF qualities that we investigated is the provision of biodiversity benefits via high quality artificial reef habitat. Respondents also prefer siting OWFs further from shore so they are less visible, and ownership that is not private.

3.3.1 Model results: strong preference for biodiversity benefits

We provide model coefficients and marginal WTP associated with going from a wind farm associated with a small biodiversity loss (30%), privately owned and 1 mile from shore to wind farms with the various characteristics (see Table 3.2 and Variables in choice experiment). The

“opt-out” fossil fuel option (Option C) was chosen 10.5% of the time while the wind farm options (Option A or B) were chosen 89.5% of the time. We report results from conditional and mixed logit models in Table 3.3.

63 Table 3.3. Choice experiment conditional and mixed logit models with WTP estimates (N=400). The base case was small loss of biodiversity, privately owned and 1 mile from shore. Conditional Mixed Logit Logit Variable Estimate Std. Error WTP($) Odds Ratio Estimate Std. Error WTP($) Odds Ratio big.loss -1.494 *** 0.096 -21.21 0.224 -4.153 *** 0.408 -20.29 0.016 small.gain 1.556 *** 0.081 22.09 4.741 4.338 *** 0.360 21.19 76.577 big.gain 2.416 *** 0.103 34.30 11.198 6.981 *** 0.539 34.10 1075.466 municipal 0.368 *** 0.086 5.22 1.445 1.253 *** 0.226 6.12 3.501 state 0.416 *** 0.085 5.90 1.515 1.173 *** 0.227 5.73 3.230 cooperative 0.164 . 0.097 2.33 1.178 1.603 *** 0.354 7.83 4.968 mi4 0.334 *** 0.090 4.74 1.396 0.981 *** 0.245 4.79 2.668 mi8 0.463 *** 0.088 6.57 1.589 1.309 *** 0.220 6.39 3.702 mi10 0.968 *** 0.120 13.74 2.633 2.095 *** 0.345 10.23 8.123 bill -0.070 *** 0.006 0.932 -0.205 *** 0.019 0.815

Log-Likelihood -2012 -1492.4 McFadden R^2 0.33993 0.510 AIC 4047.9 3118.8 With demographic variables Variable Estimate Std. Error WTP($) Odds Ratio Estimate Std. Error WTP($) Odds Ratio big.loss -1.494 *** 0.096 -21.14 0.224 -4.168 *** 0.385 -20.44 0.015 small.gain 1.561 *** 0.081 22.08 4.765 4.554 *** 0.347 22.33 94.999 big.gain 2.455 *** 0.366 34.73 11.652 8.514 *** 1.052 41.74 4982.165 municipal 0.371 *** 0.086 5.24 1.449 1.088 *** 0.235 5.34 2.969 state 0.420 *** 0.085 5.94 1.522 0.996 *** 0.241 4.88 2.707 cooperative 0.171 . 0.098 2.41 1.186 1.530 *** 0.353 7.50 4.617 mi4 0.337 *** 0.091 4.77 1.401 0.504 * 0.240 2.47 1.656 mi8 0.466 *** 0.088 6.59 1.593 0.973 *** 0.235 4.77 2.645 mi10 0.978 *** 0.120 13.84 2.660 1.660 *** 0.351 8.14 5.261 bill -0.071 *** 0.006 0.932 -0.204 *** 0.019 0.816 big.gain:age -0.015 * 0.007 0.985 -0.044 ** 0.016 0.957 big.gain:female 0.480 ** 0.162 1.616 -0.065 0.362 0.937 big.gain:white 0.285 0.210 1.330 -0.053 0.478 0.948 big.gain:univ_degr 0.227 0.173 1.255 0.529 0.363 1.697 big.gain:income -0.035 0.033 0.965 -0.002 0.067 0.998

Log-Likelihood 2003.9 -1488.2 McFadden R^2 0.3425 0.51175 AIC 4041.787 3120.471

Indication of significance codes: *** 0.001; ** 0.01; * 0.05; . 0.1

64 All models show significant estimates (p < 0.05) for the various wind farm features (impacts to biodiversity, ownership types and distance from shore), except the conditional logit models that estimate the cooperative attribute as borderline significant (p-value greater than 0.05 but less than 0.1). Both mixed and conditional models estimate significant and negative estimates for

60% reduction in biodiversity (big.loss), meaning there is a strong preference not to choose the wind farm that reduces biodiversity. The largest model estimates are associated with 60% increase in biodiversity (big.gain). We were most interested in demographic features that may influence the selection of the 60% increase in biodiversity so we interacted this variable with demographic variables. The negative estimates associated with the interaction between age and biodiversity gain (big.gain:age) is statistically significant (p <0.01) but the effect sizes are small in the models (-0.015 in the conditional model, -0.044 in the mixed model), indicating that older residents may be slightly less likely to choose and therefore somewhat less willing to pay for large biodiversity gains. The conditional model, but not the mixed logit model, found that women were somewhat more likely to choose the farm with large biodiversity gains (the estimate was 0.480, p < 0.01). We found no evidence that other demographic characteristics influence the selection of wind farms with 60% biodiversity gains.

3.3.2 Estimates of willingness to pay for offshore wind farm characteristics

Our M-Turk sample is younger, somewhat more female, more educated, and has lower household income than these states’ populations based on census data. Gender may have a significant impact on choices made, but this is not clear since only the conditional model shows gender as significant. Both conditional and mixed logit models suggest that older respondents may be less willing to pay for large biodiversity gains than younger respondents (big.gain:age

65 estimates are negative). If we assume that the coefficients are not biased by our sample, one way to account for the demographic differences between the M-Turk sample and census data is to estimate the WTP for each year of age difference between the samples. The average M-Turk age is 8 years younger than the census data average age. Based on the mixed logit with demographic variables model output, reduced WTP per year of age is ~$0.21 (big.gain:age estimate of 0.044 divided by the bill estimate of 0.204 is $0.216), so the 8 year difference could be estimated as reducing WTP by ~$1.73 ($0.216 multiplied by 8). The other demographic variables were not significant in the best-fit model so we do not correct for other discrepancies between the M-Turk sample and census data.

The highest addition to the monthly utility bill offered in the choice experiment ($20/month) was below the predicted WTP values for wind farm attributes (e.g., $34/month for a big gain to biodiversity according to the mixed logit model with the lowest AIC). Many of our WTP values are beyond the range of the offered payment mechanism, i.e., greater than $20, so we have lower confidence in these estimates. Extrapolating to regional WTP for OWFs with a baseline of small biodiversity loss to large biodiversity gains yields an estimate of $451 million/month (~$34.10 minus $1.73/month multiplied by coastal New England population of 13,952,200) or $5.42 billion/year ($451 million multiplied by 12 months).

66 Willingness to pay for offshore wind farm attributes

0 WTP ($/month)

−20

4 miles 8 miles 10 miles municipal state coop big loss small gain big gain Offshore Wind Farm Attributes 2

Figure 3.3. Willingness to pay (WTP) for offshore wind farm attributes. WTP is estimated for how distant a wind farm is from shore, ownership type and impact on biodiversity. These results are based on the mixed logit model, which had the lowest AIC score.

3.4 Discussion

Our results show widespread support and willingness to pay for ecologically regenerative renewable energy technology, which offers a more optimistic direction for environmental research than the predominant environmental discourse which has focused on limits (Meadows et al., 1972), boundaries (Rockström et al., 2009; Steffen et al., 2015) and declines (MA, 2003).

The consequences of this arguably uninspiring emphasis on scarcity and sacrifice may be seeding and perpetuating doubt and indifference rather than active engagement when it comes to addressing environmental challenges (Gifford and Comeau, 2011; Robinson and Cole, 2014;

Shellenberger and T. Nordhaus, 2004). This emphasis on minimizing harm—making things “less

67 bad”—may simply prolong environmental degradation rather than contribute to ecological regeneration (McDonough and Braungart, 2002; Robinson and Cole, 2014).

This study provides empirical evidence that people value approaches to building renewable energy infrastructure that generate ecological abundance. The strongest driver of wind farm preference in our study was biodiversity benefit (see Figure 3.3 and Table 3.3). Our results demonstrate a significant and substantial WTP for habitat enhancement in conjunction with

OWF development when the environmental gains and losses are visually explicit.

Based on our results, New England residents may be willing to pay $34-42 more for electricity from OWFs that have marine biodiversity benefits rather than losses (60% gain as compared to a

30% reduction in species abundance and diversity). This is higher than WTP estimates identified by Börger (2015) who estimated that residents living near the Irish Sea Coast had an annual

WTP of £7 for an OWF that increased the diversity of species by 10 and £15 if species increased by 30.

It is possible that our high WTP estimates are based on respondents making snap judgments selecting the option with the graphic of the most diverse reef without fully considering the bill.

Respondents may also have overlooked the “monthly” description of the bill, despite how it was described as renewable energy fee added each month to the bill (see Appendix I. Choice experiment survey) and each choice scenario included the descriptor “addition to monthly utility bill” (see Figure 3.2). Moreover, our WTP estimates may be larger than what people would actually pay due to “hypothetical bias,” which refers to how people frequently, but not always,

68 respond with lower willingness to pay to real as compared to hypothetical valuation questions

(Carlsson et al., 2005; Cummings and Taylor, 1999; List and Gallet, 2001; Neill et al., 1994). A meta-analysis by List and Gallet (2001) suggests that respondents overstate their valuation of a good by a factor of approximately 3 when asked under hypothetical settings. We used choice based elicitation methods, a simulated voter referendum with consequences to the respondent, and we had an opt-out option, which are methods that tend to reduce hypothetical bias as compared to other methods of assessing WTP (Loomis, 2011; Murphy et al., 2005). Some studies employ “cheap talk,” which involves inserting an explicit description of hypothetical bias and why it might occur into the survey instrument prior to the WTP questions. “Cheap talk” appears to reduce or eliminate hypothetical bias (Carlsson et al., 2005; Cummings and Taylor,

1999). Other economists argue against “cheap talk” statements, on the basis that telling participants that hypothetical estimates are generally overestimates is artificially leading

(Adamowicz and Naidoo, 2016). We did not include “cheap talk” nor did we include a reminder of household monthly budget constraints. All considered, real WTP is likely lower than our estimates, and a conservative lower bound might be 1/3 of the WTP that we report.

Similar to the findings of Börger (2015), visibility of turbines had a limited and weaker influence on wind farm choice than the considerably stronger preference for farms that increase marine species diversity (see Figure 3.3 and Table 3.3). Our results align with past research demonstrating WTP to site OWFs further from shore. Studies show that people generally consider an OWF a visual disamenity (Krueger et al., 2011; Ladenburg and Dubgaard, 2007).

Danish residents’ WTP was ~$58, $121, and $153 per household per year (Euros converted to

2006 USD) for a wind farm sited 12, 18 and 50km, respectively, from the coast as compared to

69 8km (Ladenburg and Dubgaard, 2007). Krueger (2007) estimated that inland residents of

Delaware were willing to pay to $9, $13, $16, $17, $19 and $21 per month for three years to site a wind farm at 3.6, 6, 9, 12, 15 and 20 miles, respectively, away from shore as compared to 0.9 miles. Westerberg (2013), however, revealed that some types of tourists associate amenity value with an OWF at least 8km from shore, while other types of tourists only associate disamenity value with an OWF.

Our study suggests that private ownership is not preferred, which is similar to the findings of Ek and Persson (2014). Our results indicate a small but significant preference for municipally and state owned OWF rather than privately owned. There is some ambiguity related to cooperative ownership in our study in contrast to the highly significant ownership preferences in Sweden for state, municipally or cooperatively owned OWF (Ek and Persson, 2014). The conditional logit model shows cooperative ownership as non-significant, while the mixed logit results indicate it is statistically significant (~$7.50 WTP). It seems clear, however, from both models that there is support for OWFs with some degree of community or public ownership, and a WTP more for this (see Figure 3.3 and Table 3).

3.4.1 Policy implications

Our research strengthens the case for the development of ecologically regenerative offshore wind farms. This study reveals latent public support and WTP for such technology. Developing OWFs with effective artificial reefs and communicating this design feature broadly could improve public support for this renewable energy technology, which has the potential to facilitate developers obtaining consent for OWF licensing and initiating planning processes.

3.5 Conclusion

This study reveals high levels of support for the ecologically regenerative design of a type of renewable energy infrastructure. This is particularly relevant and timely as the scientific consensus on climate change has coalesced and the need to shift away from fossil fuels has become increasingly apparent. Public support for renewable energy infrastructure expansion is needed to achieve commitments made regarding carbon reduction goals and renewable energy targets. Our research provides evidence of elevated WTP for ecologically regenerative renewable energy in the form of artificial reefs associated with OWFs along coastal New England states.

Our study suggests that integrating biodiversity benefits into the design of renewable energy infrastructure could increase public support for such developments.

71 Chapter 4: Relational values resonate broadly and differently than intrinsic or instrumental values, or the New Ecological Paradigm

Sarah C. Klain*, Paige Olmsted*, Kai M.A. Chan, Terre Satterfield

*Equal lead authorship

4.1 Introduction

Conservation scientists and practitioners have often drawn on ethical constructs to articulate support for policies aimed at addressing the biodiversity crisis. To those outside the conservation community, it may come as a surprise that the “Why conserve nature?” value debate about how to motivate people to achieve conservation outcomes has become increasingly heated and arguably detrimental to conservation science despite calls for “a unified and diverse conservation ethic” (Tallis and Lubchenco, 2014, p. 27; Vucetich et al., 2015). “Traditional conservationists” advocate for focusing on the intrinsic value of nature, protecting nature for its own sake. They often focus on strategies to minimize human interference with ecological processes and invoke ethical and moral arguments to support their stance while being skeptical of corporate involvement in conservation (Soulé, 2013). Such advocates are often pitted against the “new conservationists,” who champion the instrumental value of nature, justifying and prioritizing conservation action based on nature’s benefits to people (Kareiva et al., 2012). New conservationists tend to be more open to using market-based incentives and collaborating with corporations to protect and enhance the benefits of nature to people (ecosystem services), often derived from human-dominated landscapes (Kareiva et al., 2012; Tercek and J. S. Adams, 2013).

72 Underpinning the intrinsic vs. instrumental debate is a common objective—to promote and encourage conservation actions, from the level of the individual to national governments and international decisions. Marvier (2013) and other new conservationists claim that utilitarian conservation arguments do not undermine conservation justifications based on nature’s intrinsic value or an ethical duty to protect biodiversity. Rather, many contend that instrumental arguments offer additional ethical justifications and so “potentially broaden the tent of conservation” (Marvier, 2013, p. 1). This argument aside, the instrumental-intrinsic dichotomy can be constraining or possibly alienating to many who may potentially care more and take additional action if environmental issues were framed differently (Chan et al., 2016). Reducing the importance of nature to only intrinsic or instrumental and monetized value is also not reflective of the largely intuitive ways that people make decisions and understand the world and decide what’s right (Haidt, 2007; Kahneman, 2011; Levine et al., 2015).

The burgeoning field of ecosystem services (ES)(Costanza and Kubiszewski, 2012), long associated with a purely instrumental perspective, has recently been broadened to include other perspectives on value. The ES concept became globally recognized with the Millennium

Ecosystem Assessment (MA, 2003), which emphasized diverse connections between human well-being and nature, but the category of cultural ES arguably never fit well in the publications that ensued over the next decade (Chan et al., 2012a; Daniel et al., 2012). The instrumental orientation of ecosystem services is arguably the cause of the poor fit, in part because instrumental values are by definition substitutable, whereas cultural values are often not (Chan et al., 2011; 2012b). Quantified and/or monetized ES data often omit the more intangible values that “really get at well-being” (Hannah as quoted in Chan et al., 2012a), such as connectedness

73 and belonging to a community (both human and non-human), sense of place and other culturally and psychologically mediated relationships between people and ecosystems (Russell et al.,

2013). Consequently, researchers from a wide range of backgrounds, including anthropology, political science, economics, and ecology, have begun to develop methods designed to enable social, cultural and intangible values to play a more prominent role in ES assessments and decision-making without compromising their distinct nature (Chan et al., 2012b; 2012a; Daniel et al., 2012; Gould et al., 2014; Klain and Chan, 2012; Martín-López et al., 2012; Plieninger et al., 2013). As a result of these and related efforts, the ES field is evolving to the point that the

IPBES (Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services) conceptual framework has included relational values, which are an additional conception of values, to its mandate (Diaz et al., 2015).

The hope, as argued by Chan et al. (2016), is that a relational-value framing will be more inclusive and responsive to known aspects of sources of well-being (e.g., connection to others, place attachment) than instrumental and intrinsic values, particularly when addressing how people make decisions and what they care about. In this case, we refer to framing as in the framing effect – deliberate construction of (in this case) a value statement that may influence the response. The relational “framing” is intended to present value statements such that they facilitate the connection between humans and the natural world.

Relational values encompass “eudaimonic” values — values associated with living a good life as well as reflection about how preferences and societal choices relate to notions of justice, reciprocity, care and virtue (Jax et al., 2013; Muraca, 2011; Ryan and Deci, 2001; Ryff and

74 Singer, 2008). Relational values are derived from interactions with and responsibilities to humans, non-humans, landscapes and ecosystems (Chan et al., 2016). However, despite these conceptual advances, empirical investigation has been lacking.

Here we test the application of social-ecological relational statements quantitatively, as a first step to potentially transcend the limitations of the instrumental-intrinsic dichotomy. We pilot several types of social-ecological value statements, including instrumental, intrinsic, and relational value statements as well as value statements that use metaphors to convey a value. We assess if our set of relational value statements demonstrate internal coherence as a single or multi-dimensional construct. We compare responses to relational value statements across three populations with instrumental, intrinsic and metaphorically phrased value statements.

We also address a fundamental question: How do relational values compare to other scales often used to assess strength of environmental commitment? The New Environmental Paradigm question set (Dunlap and Van Liere, 1978), subsequently revised as the New Ecological

Paradigm Scale (NEP)(Dunlap et al., 2000), is the most widely used method to measure ecological beliefs. The NEP aggregates responses to 15 (or as few as 5) statements to assess ecological attitudes, many of which address which people possess ecocentric as opposed to anthropocentric beliefs. Social scientists have used the NEP scale with diverse populations and responses have demonstrated variation along the ecocentric-anthropocentric continuum

(Nordlund and Garvill, 2002).

75 Although global values surveys using NEP show variation, the overwhelming majority of people are indeed concerned about the natural world and prefer the idea of “co-existing” with nature rather than dominating it (Nordlund and Garvill, 2002). The NEP largely aligns with an ecocentric vs. anthropocentric framing, by assessing the extent to which people recognize 1) ecological limitations to growth; 2) the importance of maintaining a balance of nature; and 3) rejection of the idea that nature “exists primarily for human use” (Dunlap, 2008, p. 6). Thus, the question remains: does the addition of relational value items add something to the study of environmental beliefs or values, perhaps complementing the NEP by offering a different framing?

An additional question is whether relational values, once tested, are instructive in explaining pro- environmental attitudes when compared to other values. The example used here involves attitudes toward a renewable energy technology. Specifically, we focus on offshore wind turbines, which have local environmental impacts and global climate benefits. Diverse and often conflicting environmental values come into play when considering if and where to build renewable energy infrastructure. The “green-on-green” debate in wind farm literature is based on conflict over the extent to which stakeholders prioritize local environmental impacts (e.g., bird strikes from wind turbines, aesthetic degradation of landscape) as compared to global environmental concern (i.e., climate change, the need to reduce carbon emissions) (Warren et al.,

2005). We evaluate the relationship between NEP scores as well as responses to relational, instrumental and intrinsic value prompts with attitudes towards building this technology.

76 This discussion of both value types and their applicability can be summarized as four research questions underpinning our survey design and stated below:

1. Do various types of relational value statements correlate as a single construct? 2. Do relational value statements (including those strongly stated) resonate with (i.e., elicit agreement) amongst diverse populations? 3. Do people respond to relational value statements in a consistently different way than the New Ecological Paradigm (NEP) scale statements? 4. Do relational values and NEP scores help explain attitudes towards wind power?

In the following sections, we outline our approach to data collection and analysis, present our results, and discuss the implications for environmental research and practice.

4.2 Methods

Our methods comprised three components: diverse sampling, comparing value types, and testing in reference attitudes towards a renewable energy technology. For our sample, we targeted three populations: farmers and international tourists in Costa Rica, and residents of U.S. coastal New

England states. Our surveys included value/attitude statements followed by Likert scales to assess agreement/disagreement. Our analysis included conducting principle components analysis and factor analysis (for correlation in patterns of responses across questions and groups of questions), calculating Cronbach’s alpha (for assessing consistency in responses across questions), and running Pearson correlation tests as well as linear regressions. Each step is described in more detail below.

77 4.2.1 Survey value statements and sample

We derived a list of value statements related to the environment including NEP, instrumental, relational, intrinsic, and values conveyed using metaphors. The instrumental value statements were derived from concepts advanced in overviews of ecosystem services (MA, 2003). The NEP statements are a selection from the standardized NEP survey instrument to assess ecological worldview(Dunlap et al., 2000). The intrinsic, relational and metaphorically phrased value statements are derived from cultural ecosystem services literature (Chan et al., 2012b; Gould et al., 2014; Klain et al., 2014; Raymond et al., 2013). The metaphor statements are a rewording of four of the relational value statements. In contrast to the metaphor statements that focus on the social-ecological relationship itself, the relational value statements express the relationship as a premise for a value statement (e.g., the kin metaphor statement, kin_m, is “I think about the forest/ocean and the plants and animals in it like a family of which I am very much a part” vs. the kin relational statement, kin_r, is “Plants and animals, as part of the interdependent web of life, are like 'kin' or family to me, so how we treat them matters”).

In all three surveys, the value statements (Table 4.1) were the final section, so as not to prime responses in other areas of the otherwise different surveys. Survey takers were asked to respond to the value prompts using a 5 point Likert scale (i.e., highly disagree = 1; highly agree = 5).

Table 4.1. Value statements used in surveys. F = Costa Rican Farmers, T = Tourists at San José airport; MT = Mechanical Turk respondents. Reverse codes were used when appropriate so high scores mean pro-environmental; y = yes; n = no. Variable Category Statement Population Reverse code comm Relational There are landscapes that say something about who we F, T, MT n are as a community, a people health Relational My health or the health of my family is related one way or F, T, MT n another to the natural environment* iden Relational I have strong feelings about nature (including all plants, F, T, MT n animals, the land, etc.) these views are part of who I am and how I live my life kin Relational Plants and animals, as part of the interdependent web of F, T, MT n life, are like 'kin' or family to me, so how we treat them matters resp Relational How I manage the land, both for plants and animals and F, T n for future people, reflects my sense of responsibility to and so stewardship of the land wild Relational I often think of some wild places whose fate I care about F, T, MT n and strive to protect, even though I may never see them myself other Relational Humans have a responsibility to account for our own F, T, MT n impacts to the environment because they can harm other people abuse NEP Humans are severely abusing the environment F, T, MT n bal NEP The balance of nature is strong enough to cope with the F, T, MT y impacts of modern industrial nations bau NEP If things continue on their present course, we will soon F, T, MT n experience a major ecological catastrophe crisis NEP The so-called "ecological crisis" facing humankind has F, T, MT y been greatly exaggerated spaceshi NEP The earth is like a spaceship with very limited room and F, T, MT n p resources decade Intrinsic Humans have the right to use nature to meet our needs, MT y even if this includes impacts that will take a decade or more to recover right Intrinsic Humans have the right to use nature any way we want F, T y

79 Variable Category Statement Population Reverse code

I think about the forest/ocean and the plants and animals in it like: ** iden_m Metaphor Something I identify with so strongly that it makes me, me F, MT n kin_m Metaphor A family of which I am very much a part F, MT n other_m Metaphor A world we must care for so that any damage doesn't also F, MT n negatively affect humans who depend on it elsewhere resp_m Metaphor Beings to which we owe responsible citizenship and care F, MT n extract Instrumental Natural resource extraction is necessary for countries to F, T y (economic) develop clean Instrumental It is important to protect nature so we have clean air and F, T n (health) water loss Instrumental We can lose forests and wetlands, as long as we are F, T y (use) keeping enough for the environment to function

* This statement was reversed for the M-Turk sample: “My health, the health of my family and the health of others who I care about is not necessarily dependent on the natural environment.” We do not recommend reversed coding this prompt because we later realized it caused confusion. ** The farmer sample responded to metaphorical statements related to forest. The M-Turk sample responded to metaphorical statements related to ocean. Tourists were not presented metaphorical statements.

Our aim with the different populations and samples is not to suggest they are representative, but to compare across different populations. We targeted three populations with different methods including online and paper-based surveys.

4.2.1.1 Online survey

For the online sample, we used Amazon’s Mechanical Turk (M-Turk) system to enlist respondents, which has become a common recruitment method for experimental research

(Goodman et al., 2012; Paolacci et al., 2010). Data outputs are generally just as reliable as those acquired with traditional recruitment methods (Buhrmester et al., 2011). We attempted to minimize selection bias in our sample by describing it on M-Turk’s HIT (Human Intelligence

Tasks) list in general terms as a survey about preferences based on different text and image-

80 based descriptions, without using any language related to ecosystems. The sample was limited to

M-Turk workers who have mailing addresses in coastal New England states (Connecticut,

Maine, Massachusetts, New Hampshire or Rhode Island). We targeted this geographic area because this survey also included questions assessing attitudes to a proposed renewable energy technology suited to this region—offshore wind farms (see Klain et al. in prep). We collected self-reported demographic data from the sample to later compare it with census data to determine the extent to which this sample is representative of the population of these states. Upon survey completion, respondents were given a code to submit to the M-Turk system for payment.

Respondents were paid $1 to take the 10-15 minute survey. Given that the typical M-Turk worker is willing to complete tasks for ~$1.40/hour (Horton and Chilton, 2010), our payment was higher than the average reservation wage to expedite participant recruitment. Incomplete responses were discarded for a total of 400 M-Turk respondents.

4.2.1.2 Paper-based survey

Two paper-based surveys incorporated value statements for two distinct populations in

Guanacaste, Costa Rica. The first (n = 260) were international tourists in Costa Rica, who were randomly sampled in the Liberia Airport upon departure from the country. This airport primarily services the coastal tourist destinations and thus all international flights at this time were to the

United States or Canada. All tourists in the departure lounge (i.e. those who arrived just in time to board did not have time to participate) during the week of May 25, 2015 were asked if they had travelled in the region, and if so if they were willing to participate in a survey. They were predominantly tourists from North America (and the U.S. in particular). The second group

81 consisted of farmers in the Nicoya region (n = 253), mostly cattle ranchers, who spend a lot of time working the landscape, while also deriving their livelihoods directly from the environment.

In sum and across all three samples, we sought this diversity as we expected farmers to have a different profile with respect to their environmental values than the other two groups; but expected the international tourists to resemble the M-Turk population more closely, insofar as they both include substantial representation of middle and upper income Americans. The farmers were randomly selected from lists provided by the agricultural extension agencies in the region, and the value statements were included as part of a survey about environmental practices on the landscape more broadly.

4.2.1.3 Sampled population characteristics

Our M-Turk population was on average younger (32) than the tourist (45) or farmer populations

(58)(Table 4.2). The tourists and M-Turk samples were a majority female while the farmers were mostly male (88% male)(Table 4.2).

Table 4.2. Demographic characteristics of our three samples.

Percentage Percentage or Mean Socioeconomic or Mean of from Reference Population Characteristics Description Sample Population M-Turk

(N = 400) 2014 US Census Annual household Income income before ~$53,000* $66,200 taxes Age Years old 32 40 Female Gender 0.59 0.51 Bachelor degree Education 0.66 0.38 or higher White Caucasian race 0.83 0.82 Tourist (N = 260) Income before Income taxes ~$75,000 Age Years old ~45 Female Gender 0.63 Farmer

(N = 253) Bachelor degree Education or higher 0.15 Age Years old ~58 Female Gender 0.12

4.2.2 Statistical analysis

We assessed the discrimination or uniqueness of each value category using factor analyses and principal components analyses. Then we analyzed each using Cronbach’s alpha to test the internal consistency within value measures.

4.2.2.1 Eigenvalues and scree test

We calculated eigenvalues and created a scree plot to determine how many factors to include in our factor analysis and PCA. Eigenvalues associated with components or factors are included in

83 descending order in a scree plot. The inflection point, or ‘elbow’ at which point eigenvalues level off, demarcates components/factors to retain while subsequent components/factors are generally ignored. A common heuristic is to retain components/factors with eigenvalues > 1, which means that the component/factor accounts for as much or more variance as a single variable (A. Field et al., 2012).

4.2.2.2 Factor analysis

Our factor analysis investigated the structure of a set of variables to determine if there are clusters of correlation coefficients, which indicate latent variables, also called factors. This method derives a mathematical model from which underlying factors are estimated. Each latent variable is associated with some amount of the observed variable’s overall variance. Eigenvalues indicate the evenness in the distribution of the variances in the correlation matrix (A. Field et al.,

2012, p. 713). They measure the amount of the variance of the observed variables that a factor explains. If a factor has an eigenvalue ≥1, then it explains more variance than a single observed variable. In general, the factors explaining the least amount of variance are ignored.

In Factor Analysis, the amount of common variance is estimated by calculating communality values for each variable. This is usually done by calculating the squared multiple correlation of each variable with the others. Factor analysis is mathematically more complex than Principal

Components Analysis. Guadagnoli and Velicer (1988) conducted an extensive literature review and found that, in general, results from PCA differ little from Factor Analysis. We conducted an exploratory factor analysis with the hypothesis that responses to relational value statements comprise a factor distinct from responses to NEP statements (see Figure 4.1).

4.2.2.3 Principal components analysis

We used both Factor Analysis and PCA to determine if factors/components could be identified within our dataset of responses to value prompts. Principal Components Analysis (PCA) assumes that the communality of all variables is 1. This assumption transposes the original data into constituent linear components. PCA identifies linear components in the data and how a specific variable contributes to the component. Factors (called components in PCA) with large eigenvalues are retained while those with small eigenvalues are ignored (see Table 4.3).

4.2.2.4 Consistency measure: Cronbach’s alpha

We calculated Cronbach’s alpha for all of our social-ecological statements to determine the extent to which responses are consistent across NEP statements and relational statements.

Cronbach’s (1951) method is loosely understood as splitting a dataset in two in every possible way, then computing the correlation coefficient for each split. Cronbach’s alpha (!)—the arithmetic average of these pairwise correlation coefficients within a group of questions—is the most common metric of scale reliability (A. Field et al., 2012).

4.2.2.5 Correlation testing of environmental values and wind farm attitudes

We created five indices, one for each value type (NEP, relational, instrumental, intrinsic and metaphor) for the M-Turk population. We calculated indices based on the average response to these prompts about a type of environmental value because results from the factor analysis, PCA and Cronbach alpha (Table 4.5, Figure 4.1) suggested that responses to NEP and relational statements are consistent and distinct from each other and the metaphor and intrinsic value

85 statement responses also had a high level of consistency (i.e., high Cronbach’s alpha for the M-

Turk responses to these statements, see Appendix M). Despite the lower consistency in responses to the instrumental value statement, we included them for exploratory purposes.

We tested the correlation between these indices and responses to questions about attitudes towards offshore wind farms, which were also on Likert scales (see Appendix N).1 We ran linear regressions to test if demographic and environmental value responses could predict attitudes towards wind power.

4.3 Results

Our results suggest that relational value statements show internal coherence as a single dimensional construct, particularly when compared to responses to NEP prompts. We identified two factors and components when NEP and relational value statements were pooled and analyzed from our three populations using eigenvalues, a scree test, factor analysis and PCA.

These two types of value statements showed high levels of internal consistency based on their high Cronbach’s alpha scores. We also found positive correlations between the M-Turk population responses to environmental value statements and attitudes towards wind farms.

4.3.1 Two distinct factors based on eigenvalues and scree test

In order to understand distinctiveness in responses to types of environmental values and determine a reasonable number of factors/components to retain in our factor analysis, we calculated eigenvalues and conducted a scree test (See Appendix K and Table 4.3) and Principal

1 In the M-Turk sample, the Likert scale used to assess environmental value phrased with metaphors had slightly different meanings than the scale use for the other environmental values (See Appendix E). In future applications of the metaphor value prompts, “3” should correspond to a neutral, not somewhat positive attitude. 86 Components Analysis (PCA)(See Table 4.4 and Appendix L). Our scree plot, parallel analysis and optimal coordinates indicate that two factors ought to be retained for the factor analysis. The acceleration factor identifies where the slope of the curve changes most abruptly, which in our data, is directly after the first factor (see Appendix K).

4.3.2 Factor analysis results: NEP is distinct from relational value

Our exploratory factor analysis shows that survey takers responded differently to relational value prompts than NEP statements (Table 4.3 and Figure 4.1). The proportion of variation attributed to Factor 1, the “Relational” Factor (0.24), is higher than the proportion attributed to Factor 2,

“NEP” factor (0.21).

Table 4.3. Factor Weights

Factor 1 Factor 2 Variable Relational NEP comm_rel 0.54 wild_rel 0.61 iden_rel 0.78 kin_rel 0.75 other_rel 0.52 0.35 abuse_nep 0.31 0.68 bal_r_nep 0.5 spaceship_nep 0.67 bau_nep 0.36 0.78 crisis_r_nep

Factor 1 Factor 2 Relational NEP Eigenvalues/SS 2.43 2.11 loadings Proportion 0.24 0.21 Variation Cumulative 0.24 0.45 Variation

Figure 4.1. Graphical results of Factor Analysis.

Our factor analysis results show a grouping of the relational questions that is distinct from the

NEP statements. The crisis NEP statement is an outlier in the pooled data (Figure 4.1), which is discussed in greater detail in the discussion.

4.3.3 Principal components analysis: NEP is distinct from relational values

A Principal Components Analysis was used to demonstrate how relational statements group together as a separate factor from NEP statements, with the NEP vectors going in different directions from the relational vectors (see Appendix L).

Table 4.4. PCA loadings based on correlation matrix. PC is principle component, h2 is communality (variance shared with other variables, which is equivalent to the sum of squares of common factor loading for a variable). PC1 PC2 h2 u2 com abuse_nep 0.32 0.73 0.64 0.36 1.4 bal_r_nep 0.17 0.7 0.52 0.48 1.1 crisis_r_nep -0.09 0.49 0.25 0.75 1.1 spaceship_nep 0.32 0.68 0.56 0.44 1.4 bau_nep 0.4 0.75 0.72 0.28 1.5 comm_rel 0.69 0.13 0.49 0.51 1.1 wild_rel 0.73 0.17 0.56 0.44 1.1 iden_rel 0.81 0.14 0.67 0.33 1.1 kin_rel 0.77 0.13 0.61 0.39 1.1 other_rel 0.62 0.34 0.5 0.5 1.5

PC1 PC2 SS loadings 3.04 2.49 Proportion Variation 0.3 0.25 Cumulative Variation 0.3 0.55 Proportion Explained 0.55 0.45

4.3.4 High levels of agreement and consistency with types of environmental value

statements

Strong relational value statements resonate with diverse populations based on how the average

response to relational value and NEP statements was 4 (Agree). The responses to NEP

statements, on average, reflect relatively high ecological concern (see Table 4.5). NEP responses

were consistent (Tourist ! = 0.79 and M-Turk ! = 0.84), except for Costa Rican farmers (! =

0.35), largely due to the farmers’ wide variation in response to the “crisis” prompt (The so-called

"ecological crisis" facing humankind has been greatly exaggerated, see Table 4.1). We did not

include instrumental or intrinsic value statements when calculating ! because of the limited

number of statements in these categories.

Table 4.5. Cronbach’s alpha, mean response and standard deviation of responses across value statements.

Cronbach’s Standard

alpha Mean deviation NEP (5) Full dataset 0.73 4.0 0.75 Farmers 0.35 4.3 0.49 Tourists 0.79 3.7 0.81 M-Turk 0.84 4.0 0.74 Relational (6) Full dataset 0.80 4.0 0.68 Farmers 0.73 4.4 0.43 Tourists 0.79 3.9 0.75 M-Turk 0.79 3.9 0.61

Costa Rican Farmers responded differently to our value statements than the M-Turk and Tourist samples. The Farmers on average responded with higher levels of agreement to relational value prompts (mean = 4.4) as compared to Tourists (mean = 3.9) and M-Turk workers (mean =

3.9)(Table 4.5). Farmers on average scored higher on the NEP scale (mean = 4.33) than Tourists

(mean = 3.65) and M-Turk workers (mean = 3.96) (Table 4.5, Figure 4.2, Figure 4.3). The relational and NEP statements as well as the distribution of Likert-scale responses across the three populations is shown in the histograms in Figure 4.3. The x-axis is the number of respondents and the y-axis is the items of the Likert scale (1 means strongly disagree to 5 meaning strongly agree).

90 Social Ecological Relational Value Statements

How I manage There are I often think of I have strong Plants and My health, the Humans have the land, both landscapes some wild feelings about animals, as health of my a responsibility for plants and that say places whose nature part of the family and the to account for animals and something fate I care (including all interdependent health of our own impacts to the for future about who we about and plants, web of life, are others who I people, reflects are as a strive to animals, the care about is environment like 'kin' or because they my sense of community, a protect, even land, etc.) family to me, dependent on responsibility can harm other people though I may these views so how we the natural to and so people never see are part of treat them environment.* stewardship for them myself who I am and matters land how I live my

life

resp_rel comm_rel wild_rel iden_rel kin_rel health_rel2 other_rel

4 Farmer 3 2 1

5 sub_pop

4 M − Farmer 3 Turk M−Turk 2 Response 1 Tourist

4 Tourist 3 2 1 = Strongly Disagree; 5 Agree 1

0 50 100 150 200 0 50 100 150 200 0 50 100 150 200 0 50 100 150 200 0 50 100 150 200 0 50 100 150 200 0 50 100 150 200 count New Ecological Paradigm Statements

Humans are severely The balance of nature The so-called The earth is like a If things continue on abusing the is strong enough to "ecological crisis" spaceship with very their present course, environment cope with the impacts facing human kind limited room and we will soon of modern industrial has been greatly resources experience a major nations exaggerated ecological catastrophe abuse_nep bal_r_nep crisis_r_nep spaceship_nep bau_nep

4 Farmer 3 2 1

5 sub_pop

4 M − Farmer 3 Turk M−Turk 2 Response 1 Tourist

4 Tourist 3 2 1 = Strongly Disagree; 5 Agree 1

0 50 100 150 200 0 50 100 150 200 0 50 100 150 200 0 50 100 150 200 0 50 100 150 200 count

Figure 4.2. Mean and distribution of responses to relational value prompts and New Ecological Paradigm Statements. The sample includes Costa Rican farmers (n = 253), tourists in Costa Rica (n = 260) and US M-Turk workers (n = 400). *The health_rel prompt for the M-Turk population was worded “My health, the health of my family and the health of others who I care about is not necessarily dependent on the natural environment.” Scores were reversed for this population when included in the analysis.

91 A shown in Figure 4.3, the M-Turk and tourist populations responded similarly to the

instrumental value statements (the standard errors overlap for 2 out of 3 instrumental value

prompts). Costa Rican farmers agreed more strongly with the metaphorical statements than the

M-Turk population. Except for the “crisis” statement, Costa Rican farmers scored the highest on

the NEP scale, followed by M-Turk then the Tourist population. The M-Turk and Tourist

populations responded similarly to the relational value prompts and lower than the farmers

(except for the similar responses to the responsibility prompt, “resp_rel”).

Farmer 3 M−Turk

Response Tourist 4 = Agree; 5 Strongly Agree 2

1 = Strongly Disagree 2 Disagree; 3 Neither Agree nor 1

Extract Loss Clean Kin Resp Inden Other Decade Right Abuse Bal Crisis Space Bau Comm Wild Resp Iden Kin Health Other met met met ship rel rel rel

Instrumental Metaphor New Ecological Paradigm Relational Intrinsic

Figure 4.3. Mean response with standard errors to value prompts across three populations. Red circles indicate the mean response across the populations for each value statement.

Out of all of the environmental value statements that we tested, the highest average response for

the M-Turk and Tourist population was agreement with an instrumental value: It is important to

92 protect nature so we have clean air and water (“Clean”)(. Two NEP statements (“BAU” and

Mean Response to ValueAbuse”) Prompts ranked highest for the farmer population as shown in Figure 4.3 and Table 4.6. 5 4

Farmer 3 2 Table 4.6. Top six mean responses to environmental value statements across three populations. 1 The top four farmer scores are not statistically different from each other, effectively all being tied for first, Type of 0 comm_rel is statistically different from the first two, bau_nep and abuse_nep. 5 Rank M-Turk Tourist Farmer Value Prompt 4 1 Clean (4.69) Clean (4.6) BAU (4.81) Instrumental M

3 − 2 Other (4.34) Other (4.4) Abuse (4.81) Turk Intrinsic 2 3 Abuse (4.25) Responsibility (4.3) Other (4.75) Metaphor 1 Response 4 Other (4.09) Right (4.1) Spaceship (4.74) 0 NEP 5 Community (4.07) Community (4.1) Community (4.70) 5 Relational 4 6 Right (4.00) Health (3.9) Responsibility (4.58) Tourist 3 2 1 = Strongly Disagree; 2 1 0 clean other resp abuse other_m comm right kin resp_m crisis bau spaceship4.3.5decade Majorityiden of balM-Turkloss samplehealth havewild positivekin_m attitudesiden_m extracttowards wind farms Social−Ecological Value Prompt

3 = Neither Agree nor Disagree;4 Agree; 5 Strongly The majority of our M-Turk survey takers had positive attitudes towards wind power both at a

national and state level (see Figure 4.4).

What is your attitude toward developing In your opinion, construction of offshore wind power in the US? wind turbines off the coast of your state should be:

60%

40% 40%

20% 20%

0% 0%

Positive Neutral Negative Not sure Tolerated Prohibited Very positive Very Negative Encouraged Discouraged Attitude Opinion

Figure 4.4. Attitude toward wind at the national (left) and state level (right).

93 As shown in Figure 4.5, a total of 77% of respondents thought that an offshore wind farm would have no difference on if they went to the coast for recreation, 14% said less likely or much less likely, while 10% said more likely or much more likely. Other responses to wind farm attitude questions are reported in Appendix O.

Would the presence of a visible offshore wind farm make you more or less likely to go to the coast for recreational purposes, e.g., beach−going, boating, fishing, or walking along the coast? 80%

60%

40%

20%

Less likely More likely No difference Much less likely Much more likely

Figure 4.5. Expected impact of an offshore wind on going to the coast for recreation.

4.3.6 Significant correlations between wind farm attitudes and environmental values

We calculated Pearson’s r correlation coefficients between indices comprised of the mean responses to NEP, relational, instrumental and intrinsic value prompts and attitudes towards wind farms using. See Appendix N for explanations of variables.

94 att_w_US

0.50 const_st ***

0.21 0.26 wf_rec *** ***

0.17 0.02 − coast_rec **

0.30 0.27 0.38 − first_st *** *** ***

0.11 0.07 0.05 0.20 − oper * ***

0.36 0.20 0.09 0.04 0.24 0.05 NEP *** *** ***

0.26 0.21 0.06 0.18 0.28 0.05 0.60 relational *** *** *** *** ***

0.26 0.13 0.06 0.22 0.19 0.02 0.45 0.71 metaphor *** * *** *** *** ***

0.17 0.12 − 0.11 0.14 − 0.63 0.57 0.40 intrinsic *** * * * *** *** ***

0.20 0.06 0.09 0.11 0.16 0.00 0.54 0.47 0.36 0.57 instrumental *** * ** *** *** *** ***

Figure 4.6 Correlation matrix of attitudes towards wind farms and environmental values. Red denotes a negative correlation while blue is positive. P-value of < 0.0005 is "***", <0.005 is "**", <0.05 is "*".

As shown in Figure 4.6, the five types of environmental value indices positively correlate with attitudes towards developing wind power in the US (p<0.0005) and attitudes towards supporting a wind farm in your state if it was the first of many (p ranges from >0.05 to 0.0005). The correlation is positive and significant between NEP, relational, metaphor and intrinsic value indices and support for turbine construction along a respondent’s state’s coast (p ranges from

<0.0005 to <0.05). The five value indices positively correlate with each other (p <0.0005).

Additionally, the five value indices positively correlated with frequency of recreating on the coast (p ranges from <0.005 to <0.005).

4.3.7 Environmental values influence wind farm attitudes at national and state level

We created three simple linear models for fixed effects to predict attitudes towards wind power in the M-Turk population. Dependent variables of wind farm attitudes were predicted based on indices of four types of environmental values (NEP, relational, metaphor, instrinsic and instrumental) as well as demographic characteristics (gender, age, education level, and income).

Significant regression equations were found: 1) for wind power in the US (F(9, 390) = 9.771, p<0.001), with an R2 of 0.165; 2) construction of a wind farm off a respondent’s states’ coast

(F(9, 372) = 3.040, p<0.001), with an R2 of 0.046; and 3) support of a wind farm in a respondent’s state if it was the first of many (F(9, 390) = 5.357 p<0.001), with an R2 of 0.089.

See Table 4.7.

96 Table 4.7 Linear model results for fixed effects on attitudes towards wind power as anticipated by responses to environmental value statements and demographic characteristics. Dependent variable:

Wind power in the US | Wind farm off your state's coast | Support if first of many

(1) (2) (3)

NEP 0.286*** 0.118** 0.194** (0.060) (0.054) (0.085) rel ational 0.051 0.186 ** 0.381 *** (0.083) (0.074) (0.117) met aphor 0.084 * -0.032 -0.053 (0.050) (0.046) (0.071) intrinsic -0.053 -0.007 -0.092 (0.049) (0.043) (0.069) instr umental 0.057 -0.071 0.059 (0.064) (0.056) (0.090) gender 0.194 *** 0.078 0.125 (0.066) (0.057) (0.093) age -0.005 -0.0001 -0.003 (0.003) (0.002) (0.004) edu cation -0.025 -0.011 -0.054 (0.024) (0.022) (0.035) income 0.030 ** 0.014 -0.010 (0.013) (0.011) (0.018)

Constant 2.727 *** 2.828 *** 1.950 *** (0.318) (0.281) (0.449)

Observations 400 382 400 R2 0.184 0.069 0.110 Adjusted R2 0.165 0.046 0.089 Residual Std. Error 0.610 (df = 390) 0.521 (df = 372) 0.861 (df = 390) F Statistic 9.771*** (df = 9; 390) 3.040*** (df = 9; 372) 5.357*** (df = 9; 390)

Note: *p <0.1, **p<0.05***p<0.01

We found significant positive correlations between all of our environmental value indices and attitudes towards wind power at a national level, offshore wind farms at a state level and support for a wind farm if respondents knew it was the first of many (see Figure 4.6). The linear models

(see Table 4.7) suggest that different types of values play stronger and weaker roles in influencing attitudes towards wind farms at different scales. NEP scores have a larger influence

97 on attitudes towards wind power at a national level than the other types of environmental value.

Both NEP scores and relational values influence attitudes towards wind at a state level.

Relational values appear to play a stronger role than NEP scores in influencing attitudes towards supporting a wind farm if respondents are told it was the first of many (see Table 4.7).

4.4 Discussion

This research is a first step in seeking to operationalize a “relational values” construct in a survey form in reference to other widely used constructs (intrinsic and instrumental) and a measure of environmental concern (NEP). The following sections discuss the research questions in turn. The first is associated with the relational concept in general, namely that diverse populations agree with the statements, suggesting that what we refer to as a “relational framing” (in terms of the phrasing rather than as an experimental design) is widely resonant. The following two sections discuss how responses differed between the relational statements and the NEP, followed by how there was consistency in responses to the relational statements, which could lead to treating this set of statements as an index. Also, the correlations between wind farm attitudes and positive relational and NEP responses, theoretical and policy implications of these findings and proposed paths forward are discussed.

4.4.1 Diverse populations tend to agree with strong relational value statements

Agreement with relational values was higher than anticipated across populations. The mean response for all three of the populations to the relational value statements was 4 (see Table 4.5,

Figure 4.2 and Figure 4.3), which is equivalent to “agree” on the Likert scale. The average for each relational value prompt differentiated by population was higher than 3.6. We had expected somewhat lower means given the explicit nature of the social-ecological linkage and our

98 deliberate attempt to phrase the prompts strongly to foster variation in our sample. The relational prompts therefore push the bounds of how people think about the environment in relation to themselves – such as thinking of wildlife as kin and considering the environment as part of their identity. Although environmentalism may have become marginalized in the last decade (Marvier and Wong, 2012, p. 292), these social-ecological relational statements clearly resonate with our

M-Turk, tourist and farmer samples (i.e., respondents tend to agree and strongly agree with the value statements) (Figure 4.2).

The comparison between the relational value and metaphor statements is instructive, suggesting that although social-ecological relations are lower in North American populations, associated values remain strong in the populations we surveyed. M-Turk samples tend to be comprised of

~90% urban residents (Huff and Tingley, 2015). The farmers’ responses to the metaphor statements were significantly higher than the M-Turk responses, and in the same range as their relational responses. The M-Turk population responses to the metaphorical statements were significantly lower than both the farmers and the M-Turk relational responses (Figure 4.3). We speculate that the farmers are comfortable talking about nature in a deeply relational way, while the M-Turk population is likely less comfortable with such ‘relationality’, but can still agree with the moral conclusion expressed in the relational statements. We view this as further indication that a relational framing may be an accessible way to engage diverse parties for the purpose of conservation, including those who do not have an ecocentric worldview.

Relational value responses do not have the highest average among the types of value statements in the three populations (Table 4.6). Out of the 17 statements presented to all three populations,

99 the overall highest ranked statements (in two of the three populations, tourist and M-Turk) was the “clean” statement: “It is important to protect nature so we can have clean air and water.” We classified “clean” as an instrumental statement (Table 4.1), but it is not narrowly self-oriented, in that it implicitly includes concern for the well-being of others. The highest overall statement for the farmers was “bau” (“If things continue on their present course, we will soon experience a major ecological catastrophe,” i.e., business as usual). However since the farmers were so high in their responses overall—their top 5 responses averaged over 4.7, meaning that the majority of respondents answered 5— the differences between the top 5 are not significant (with the exception of the fifth being different from the first and second rank based on t-test results—

Table 4.6), thus the top four could all be considered a top response.

It is not surprising that relational values were not noticeably higher in the farmer population as compared to their NEP scores. We perceive the benefit of relational values is that it may allow people to express environmental concern that they otherwise would not (on a scale like the NEP, for example). For people with already high environmental values, it is not surprising they score equally high in this alternative framing.

The top six overall mean scores of our three populations are depicted in Figure 4.3. For the tourist population, four of the top six mean scores were relational statements. All three populations included the “community” statement as the fifth highest. The M-Turk and farmer population shared two of the top five (“community” and “other”). The community statement refers to recognizing the uniqueness associated with place, where as “other” refers to responsibility to reduce environmental harms felt by people elsewhere. All six relational

100 statements are represented in the top 6 value statements when all three populations are combined, suggesting 1) there is resonance of relational statements in general, and 2) different aspects of relational values resonate with different populations, that is, averaging across different populations we see high levels of agreement with several relational statements.

4.4.2 Relational value responses are distinct from NEP

The factor analysis (FA) and PCA tests (Table 4.3, Figure 4.1, Table 4.4) reveal a distinction between relational value responses and the NEP. Additionally, this analysis allows comparison across statements and sets of question to determine the consistency with which individuals and subpopulations responded to the survey, enabling underlying factors to emerge (Child, 1970).

The statements cluster in the factor analysis differently as individual populations (see Appendix

J) as compared to pooled results (Figure 4.1), but in all four cases the distinction between the two sets is clear. Examining uniqueness of the relational statements as compared to the NEP, the former has a higher proportional variation in the pooled data set (Figure 4.1), meaning the relational statements are more tightly knit as a group than the NEP.

The PCA reveals two principle components, which consist of sets of variables that correlate with each other. This can be seen in the trajectories of the vectors in the graphical results of the PCA

(Appendix L) and the weightings of PC1 and PC2 columns (Table 4.4). Both the PCA and factor analysis are used for similar objectives but make different assumptions (see Methods section).

Both demonstrate that the relational statements fall into distinct components or factors, which supports the hypothesis that the relational framings induce a different but coherent response

101 pattern. This response is also consistent, as evidenced by the high α across the relational statements (Table 4.5).

4.4.3 Relational statements can be a single construct and have potential as new index

Our Cronbach’s alpha scores suggest, somewhat to our surprise, that the six relational values statements cluster together strongly as an index. The six statements capture different aspects of values about relationships with nature, and are not intended as multiple expressions of the same idea, so it is interesting how strongly the statements do cluster. This result was echoed in the tourist and M-Turk population, with α scores of 0.79 and 0.84 respectively, whereas the farmers had a score of 0.35. The exception driving this unexpected result is the farmer response to the crisis statement; the widely distributed spread of responses for this statement can be seen in

Figure 4.2.

Typically, the expectation is that those with a tendency toward an ecocentric worldview will score low for this statement (until it is reversed for the purpose of analysis), and those with anthropocentric worldview will score highly. The farmer results across all statements (see Figure

4.2 and Figure 4.3) demonstrate consistently high mean responses that are also statistically higher than the other two populations as noted by the t-test results. This rural population of predominantly small-holder Costa Rican farmers are reliant upon environmental conditions for their livelihoods, and thus their strong environmental values (as understood through all of their responses) are expected. This is reflected in their high scores, and in the case of the abuse statement, statements where not a single farmer answered lower than a 4 (i.e. all respondents answered agree or strongly agree). This brings in the question of why the farmers did not follow

102 the pattern of eco-centrism, which is associated with strong environmental values and evident here.

We propose two possible explanations for the anomaly, but do not believe this is problematic for our overall results. The first possibility is wording. The statement reads, “the environmental crisis is greatly exaggerated,” with the expectation that those answering 4 or 5 (agree or strongly agree) are not as concerned about the environment as 1 or 2 (strongly disagree or disagree). It is conceivable in this region that those answering highly are deeply concerned about environmental issues, but it is such a focal point that from their perspective it is overemphasized. That is, their agreement with the statement speaks to the strong wording of “great exaggeration” rather than suggesting environmental issues in their region are not present. An additional possibility is that these farmers are better equipped to cope with change than their neighbours, thus reducing an overall sense of urgency. All farmers who responded 4 or 5 to this question (about 30%) responded in the expected NEP pattern matching an ecocentric worldview, so we do not believe that our subset of farmers lack ecocentric views. In any case, this result did not impact the analysis dramatically insofar as the NEP and relational factor analyses remained separate across all populations and as demonstrated in Table 4.3 and Figure 4.1.

Farmer anomaly aside, the inclusion of NEP statements enabled us to demonstrate that for the most part the statements correlated as expected, and our populations behaved consistently with

NEP experiments elsewhere. The high Cronbach’s alpha scores across the individual populations and all three pooled means people responded consistently to the NEP and social-ecological relational statements. In general, an alpha of 0.7 and higher is considered strong (Mohsen

103 Tavakol, 2011). Our high relational value alpha of 0.8 suggests there may be potential to generate a scale or index when considered collectively as a group, and we consider the development of such an index an avenue for future research.

4.4.4 Theory implications

As proposed in the introduction, we see potential to utilize relational values as a means to solidify or enhance connections to the natural world, by invoking other held values that are not necessarily environmental. That is, instead of thinking of nature as external or “outside of oneself,” by connection to family, places we care about, and human well-being, ‘nature’ becomes part of an individual’s realm of care.

We refer to relational values as a framing rather than as a novel way of thinking about the environment to recognize and emphasize that we are not suggesting this is entirely new conceptual territory. Environmental values have been studied extensively, along with their connections to attitudes and behaviours (Stern et al., 1995, Dietz et al., 2005, Spash et al., 2009).

Likewise, the attributes captured by our value statements were selected based on existing studies and theory that suggest associations with family, community, and identity are powerful and meaningful ideas that people will take action to protect and uphold (Martín-López et al., 2007;

Nichols, 2014). Our eventual aim is to examine whether this new value-frame can augment and support existing theories of value that posit pathways between different categories of values (and beliefs in the NEP sense of the word) and behaviour. This study is not sufficient to do so, but our data does point to some encouraging possibilities for continuing along this path. Here we discuss

104 how we envision the relational framing to contribute to the values, beliefs and norms framework

(Dietz et al., 2005; Stern et al., 1999).

Values, beliefs and norms (VBN) theory of concern for the environment suggests that there are relationships linking 1) the acceptance of basic values; 2) believing that something important is threatened; and 3) the activation of a personal norm (obligation) to take action to restore those values (Dietz et al., 2005; Stern et al., 1999). VBN posits that values influence our worldviews, which in turn influence our beliefs of how environmental change has consequences for our values, and these beliefs underlie norms from which we take action (Dietz et al., 2005). Figure

4.7 outlines the VBN theory in green, and highlights in purple how we imagine our selected relational value dimensions contribute to this pathway. Our results are far too limited and preliminary to support the hypothesis that social-ecological relational framing influences behavioural intention (let alone behavior —even the VBN theory does not claim to comprehensively explain pro-environmental behaviour), but we propose future studies to test this.

105

Figure 4.7. Value-belief norm model (green) with our proposed ways in which relational framings (purple) could influence steps of this pathway (black dashes).

We acknowledge the variety of barriers between behavioral intention and pro-environment behavior (dashed blue line).

Figure 4.7 highlights where our relational value framings might support the theorized linkages to the VBN. We propose that by leveraging some of the components of the model—namely responsibility to others (both human and non-human) and personal norms—the pathway may be strengthened or some of the other components may be bypassed. For example, a mother with anthropocentric views and little understanding of consequences of a particular threat where she lives (such as climate change influencing flooding), may still be induced to support a new coastal

106 protected area in her community, if doing so is consistent with notions of good parenthood or citizenship.

Reflecting upon our results in the context of this diagram, we note that the highest scores from the relational statements were those that referred to groups in which they are a part or to which they feel a sense of responsibility, including family and community. Psychological evidence points to the importance of in-groups, social norms, and peer-pressure to influence behavior, both in general and with pro-environmental behaviours specifically (Cialdini and Goldstein,

2004; Crompton and Kasser, 2010). While instrumental and intrinsic values tend to focus on individual ways of thinking about the world, we propose relational framings have the capacity to establish or enhance social influences that encourage action.

4.4.5 Policy and practical implications

Governments, NGOs, and decision-making bodies wrestle with how to effectively engage communities in environmental decision-making processes (Reed, 2008). Regulatory bodies and environmental impact assessment require consultation, yet assessments tend to focus on biophysical impacts and have struggled to capture cultural ecosystem services, due to their less tangible and less quantifiable nature. We propose there is a gap in the traditional tools that explore and explain values on how we relate to the environment. Relational values may be used to frame or facilitate discussions in decision-making processes linking environmental change to tangible and intangible values. Here again we refer to framing in terms of a value construct, rather than comparative framing used in experimental designs. Methods to assess social- ecological relational value could be further refined to characterize how communities or

107 individuals think about the environment. Invoking relational values may be key to reframing conservation policy approaches (Berbés-Blázquez et al., 2016).

Framing conservation with relational values may offer more powerful leverage for conservation than emphasis on instrumental or intrinsic values. Intrinsic values in and of themselves are enough to motivate only a minority of people to achieve conservation goals (Armsworth et al.,

2007). A potentially broader array of people can be motivated by appeals to financial benefit and self-interest in the name of conservation, but such appeals reinforce ‘extrinsic’ values—those associated with the pursuit of prestige, power, image and status. Psychological research has shown that reinforcement of extrinsic values can suppress intrinsic values, which are linked to concern for others and the environment, kindness, understanding, appreciation, tolerance and protection of people and nature (Blackmore et al., 2013). Furthermore, an instrumental-value basis for conservation can only motivate conservation that is demonstrably useful (Chan et al.,

2007).

Relational value statements could be a part of how the International Union for Conservation of

Nature’s Key Biodiversity Areas partnership conducts biodiversity documentation, which would include consistently collected information that assists policy advocacy on-site, as well as broader analysis to prioritize areas for conservation. This partnership, as just one example of a potential application of relational values, identifies important sites for various taxa, and is currently consolidating a variety of partners to create a framework for assessment (threats, associated ecosystem services, etc.) (Eken et al., 2004). These data could support prioritizing conservation actions and policies that resonate with people locally. In a similar vein, diverse

108 conceptualizations of values are incorporated in the conceptual framework of the International

Panel for Biodiversity and Ecosystem Services (IPBES). Relational value statements may help operationalize these diverse conceptualizations in the planned regional assessments.

We anticipate the concern that employing community values or framing options could be used to merely leverage instrumental values. Though we do not explicitly test that, our hypothesis relates to encouraging environmental values in those who may not already feel strongly by anchoring them to something they already care about and with which they already identify (e.g., community, family).

Our results linking environmental values to attitudes towards wind power and offshore wind farms suggests that strong social-ecological relational value may influence support for a wind farm at a state level and if it is a pioneering project, leading the way for many others to come.

Relational values do not have a statistically significant impact on attitudes towards wind power at a national scale, suggesting that relational values may have more influence at a state level.

Our intention is not to find another avenue to “sell” the environment and its associated benefits to a broader audience. As highlighted by Chan et al., “To be more than mere marketing, environmental management must reflect on and possibly rethink conservation in the context of local narratives and struggles over a good life” (p. 1464).

109 4.4.6 Proposed paths forward

Our first pass at assessing social-ecological relational values resulted in a preliminary assessment scale that can help launch future research. Our objective was not to create a new, universally valid scale for social-ecological relational values. Although we capture diverse types of relational values, we do not claim to have captured all aspects of “relationality.” We acknowledge there may be different and/or additional statements that could enrich a social- ecological relational index. We can imagine several research trajectories, as well as how other future research may augment the ambitions of this preliminary study.

• Expand and refine social-ecological relational statements. Our six relational

statements are likely not comprehensive. We can imagine further dimensions to be tested,

such as the extent to which natural elements contribute to a sense of belonging. Index

development in the psychological literature entails including more overlap between

statements to probe similar themes in multiple ways and test agreement with various

statements in different cultural settings (if universality–to the degree it is possible–is the

goal). The list should be refined list until there is greater certainty of its appropriateness

and accuracy for assessing the presence and strength of social-ecological relational

values.

• Explore social-ecological relational values with other methods. Surveys can be useful,

but other methods, such as interviews and focus groups, can help delve into the

complexity and context-specific dimensions of social-ecological relational values.

110 • Use social-ecological relational value statements as an index in before/after or

control/impact studies. Such research would shed light on values in the context of

various environmental management and conservation interventions.

• Embed social-ecological relational values research in scenarios with real-world

constraints. We envision empirical testing of relational values in the context of tradeoffs

and/or external constraints, including scenarios or choices to more accurately reflect the

types of decisions people make on a daily basis. One particular set of people whose

behaviours are of particular interest includes consumer responses to relational framings,

and testing consumption behavior when the disconnect between consumption practices

and environmental impact are made more explicit.

• Further test relational value statements in comparative framing experimental

designs to estimate influence of relational values on renewable energy development

and energy conservation. Our exploratory analysis suggests that relational values may

influence attitudes towards wind farms at a state level. Future research could focus on

local levels and if relational value frames could influence support or rejection of sites for

renewable energy development. We also suggest research on the extent to which

relational value considerations could increase motivation for energy conservation if direct

connections are made between energy consumption and ecological consequences.

111 4.5 Conclusion

The study provides preliminary empirical evidence of widespread support for social-ecological relational values, an emergent topic in conservation (Berbés-Blázquez et al., 2016; Chan et al.,

2016). We foresee diverse paths forward to test this idea of relational values as a means of overcoming the instrumental vs. intrinsic value of nature debate.

Self-interest tends to prevail when instrumental values dominate communications, campaigns and debates (Blackmore et al., 2013). Instrumental values, however, are one type of the various values that can come into play when we make decisions. Insights from cognitive psychology highlight how we often make decisions and act based on affective responses to situations rather than mental calculations of utility associated with different outcomes (Kahneman, 2011; Levine et al., 2015). Similarly, while we acknowledge the logic behind instrumental justifications for biodiversity conservation, studies show numerous other values, beliefs and attitudes motivate conservation action, including, but not limited to, identity and social norms, biophilia, altruism and notions of reciprocity. Leveraging these motivators in relational terms might engage more people and enable individuals and communities to rethink conservation in the context of local narratives and what it means to pursue a good life, which goes far beyond focusing on instrumental values (Chan et al., 2016).

This study suggests a relational value framing as a new direction for innovation when it comes to ecosystem service assessments, designing conservation initiatives and potentially building support for renewable energy. This could not only inform, but also inspire the action necessary to cultivate a future better for humans and other species.

112 Chapter 5: Will communities “open-up” to offshore wind? Lessons learned from New England islands

Sarah C. Klain, Terre Satterfield, Suzanne MacDonald, Nicholas Battista, Kai M.A. Chan

Preface

This chapter was the result of Sarah Klain’s participation in the UBC Public Scholars Initiative, an innovative program to support collaborative scholarship that contributes overtly to the public good. Sarah Klain collaborated with the non-profit organization Island Institute. Together, they devised a transdisciplinary research agenda to understand successes and shortcomings across a set of community engagement efforts pertaining to proposed offshore wind farms. A major project goal was to better link academic research with civic practice and decision-making.

5.1 Introduction

The scientific consensus regarding the urgency of climate change mitigation has coalesced

(IPCC, 2014) while ideological and economic debates about appropriate actions and energy policies have become increasingly polarized (Campbell and Kay, 2014; Dunlap and McCright,

2008; Kahan et al., 2012; McCright and Dunlap, 2011). Achieving the IPCC’s goal of 1.5°C or less of warming entails a transformation of various modes of production and consumption, including massive changes in our energy infrastructure (Johansson et al., 2016). Transitioning to low carbon sources of electricity largely depends on the extent to which people act at various scales to obstruct (e.g., file lawsuits), accommodate or champion low-carbon energy technology.

113 Switching to greater reliance on renewable energy can diversify sources of energy, reduce carbon emissions, reduce air pollution and meet growing demands for electricity (Jacobson and

Delucchi, 2011). Accordingly, renewable energy infrastructure is becoming increasingly common in and near where people live. In 2015, the U.S. committed to increasing non- hydroelectric renewable energy generation to 20% of the U.S. total by 2030. This includes a projected 22,000 MW of offshore wind, which could power 4.5 million homes (DOE EIA, 2015;

OPS, 2015).

Siting offshore wind farms and other renewable energy infrastructure has often been controversial, resulting in project delays and cancelations (Kimmell and Stalenhoef, 2011;

Roberts et al., 2013). Bell et al. (2005) identified a ‘social gap’ when it comes to understanding why national opinion polls reveal high levels of public support for the development of renewable energy while specific applications for its development have low success rates. Proposed explanations for this ‘social gap’ include the following: 1) self-interested NIMBY-ism (not in my backyard), defined as “an attitude motivated by concern for the ‘common good’ and behaviour motivated by ‘self-interest’” (Bell et al., 2005, p. 460); 2) democratic deficit in that a small, unrepresentative number of opponents dominate the decision processes; 3) qualified support in that national surveys may report high levels of public support, but this support may in reality be based on certain conditions being met (e.g., related to noise, size, number of turbines, environmental protection, community engagement, fairness of decision process, and fair allocation of economic benefits); and 4) place protectors, who perceive higher place value in a specific location without the renewable energy development (e.g., rejecting a development due to its impact on local biodiversity or the historic qualities of a particular landscape), but may accept

114 the development in another location (Bell et al., 2013). If renewable energy targets are to be achieved, this “social gap” must be bridged to mitigate, accommodate or otherwise work through concerns and hostility of local communities to particular renewable energy projects (Bell et al.,

2005; Haggett, 2011).

Social science can elucidate why and how renewable energy controversies might be ameliorated via robust public engagement strategies, including those that seek to clarify both concerns and possible outcomes or alternatives. Public participation in decision-making has the potential to enhance the quality of decision outcomes while improving the capacity of those involved to meaningfully engage in policy processes (Dietz and Stern, 2008). Scholars of risk, technology and the social dimensions of renewable energy recommend shifting governance away from reliance on primarily technocratic evaluations of risks and benefits. Instead, scholars have called for methods that ‘open-up’ the capacity for people with diverse perspectives to participate in analytic deliberative processes to determine what constitutes appropriate development of a technology (Devine-Wright et al., 2011; Stirling, 2008). Analytic-deliberative methods are approaches to public engagement in decision-making that involve assessment and dialogue to reconcile technical as well as expert knowledge with citizen values (Burgess et al., 2007). Such methods can result in increased trust among those involved and acceptability of outcomes (Renn,

2008; 1999). “Opening up” decision-making processes entails recognition and accounting for the numerous factors driving the development and deployment of technology, including “individual creativity, collective ingenuity, economic priorities, cultural values, institutional interests, stakeholder negotiation, and the exercise of power” (Stirling, 2008, p. 263). And yet, when done poorly (i.e., closing down decision making), deliberative processes can ‘close’ down both

115 discussion of new technologies and so too the possibility of innovations, such as the development of offshore wind farms in North America.

Although numerous articles have been published on public opinion of offshore wind (Firestone et al., 2012; 2009), few academic studies have focused on identifying and characterizing both the successes and challenges of community engagement practices involving this technology in North

America, and how this relates to theory about analytic deliberative processes. Addressing this gap is an opportunity for social science research to inform the development of this industry and siting developments in general.

We conducted research on the experiences of three New England islands to explore both the use of deliberative designs and logics of acceptability or unacceptability of offshore wind farms. Our goal was to parse how public engagement has occurred and the types of engagement practices that built or eroded support for wind farms. We used normative theory on key components of analytic-deliberative processes to explain characteristics of community engagement that worked well versus those that resulted in relatively higher levels of frustration among various parties.

Our research identifies similarities, differences and gaps between this normative theory and our three island community contexts to identify characteristics of community engagement that may minimize frustration and increase satisfaction with decision processes and outcomes among local stakeholders.

116 5.1.1 Theorizing public engagement processes

A normative theory of public participation in decision-making has sought to conceptualize and identify principles for reaching legitimate outcomes (Figure 5.1) (Abelson et al., 2003; Renn,

1992). Concepts of ideal speech situations and communicative competence are central to this theory. An ideal speech situation involves the aspirational goal of reaching a rational consensus wherein communication follows implied rules, no coercive or non-rational pressures exist and assertions made are based on reason and evidence only (Habermas, 2004; Renn, 2008).

Communicative competence is “the ability to use language…to create understanding and agreement… This requires people enter into a discourse [i.e., discussion or deliberation exercises] with an attitude oriented toward reaching understanding. People must be committed to reflecting on their personal beliefs, values, preferences, and interests, they must be open to alternative definitions of reality, and they must listen to other people’s arguments with an open mind” (Webler, 1995, p. 44). Competence also means that the people involved in the deliberation are able to assimilate information to reach an adequate understanding of the issue and appropriate procedures are in place to choose the relevant knowledge to inform the process.

Principles of fairness are linked to competence to the extent that legitimate outcomes depend not just on competence, but fairness as concerns equality of inclusion in the decision process, procedural fairness throughout the deliberation, and mutual respect among all involved. Lastly, fairness is transgressed when 1) the role of power is ignored or is not neutralized; and/or 2) when political institutions make the deliberative process an end-creating activity, rather than the means for generating an outcome. These obstacles can block the achievement of legitimate outcomes

(Figure 5.1).

117

Equality of access

Fairness Procedural fairness Ignores or Norma)ve theory of neutralizes role public par)cipa)on in Mutual respect of power decision-making Relevant knowledge and Legi%mate Revision of Habermas’ ideal understanding of issue via Public Outcomes speech and communica%ve informa%on access and par%cipa%on competence interpreta%on can be purpose- or end-crea%ng Competence Appropriate procedures ac%vi%es used to select knowledge used to inform process

Figure 5.1. Normative theory of public participation in decision-making, adapted from Abelson et al. (2003). The meta-principles of fairness and competence are necessary (Habermas, 2004; Renn, 1992) but arguably not sufficient to reach legitimate outcomes (Ryfe, 2005). Neglecting the role of power and participation as an end unto itself rather than a means to an outcome can be barriers to reaching legitimate outcomes (Abelson et al., 2003).

Abelson et al. (2003) expand and operationalize this normative theory into pragmatic principles

for evaluating public participation in decision-making with more explicit recognition of the role

of power in deliberative processes (e.g., the availability and use of particular information can be

a source of power). This highly cited review, with over 720 citations on Google scholar as of

2016, documents how no simple formula exists for designing an optimal public engagement

process, but four key topics require attention: 1) representation; 2) procedural rules; 3)

information employed in the process and 4) the outcomes including decisions resulting from the

process. Representation refers to determining who fairly represents the “public” in a decision-

process. This can be challenging because fair and legitimate processes that provide meaningful

opportunities for learning and recognition of diverse perspectives tend to be time-intensive and

118 relatively exclusive processes that can only involve a small number of people. Further complicating fair representation is that citizens are more likely to get involved if they fear losing something they value (Abelson et al., 2003). Situations can arise when a majority of people support or feel neutral towards a proposal, but they are a “silent majority” because they choose not to get involved with the decision process (Stephenson and Lawson, 2013). Abelson et al.

(2003) documents how procedural rules can help manage this potential self-selection of who gets involved. They also identify the importance of being upfront and transparent about the timing and extent of public engagement as well as responsiveness on the part of an authority who compiles and responds to public input. Providing ample time for those involved to challenge the information presented in the process is important, as is maintaining mutual respect throughout the deliberation. Choices about information are crucial, specifically what information is selected then how it is presented and interpreted. Finally, not just the process leading to the decision, but also the outcome (the decision) needs to be associated with legitimacy and accountability

(Abelson et al., 2003).

Abelson et al. (2003) identified these key components of public participation in analytic deliberative processes based on experiences in the health sector. Numerous other studies uphold them in the design of deliberative processes related to sustainability issues (Antunes et al., 2009;

Blackstock et al., 2007; Burgess and Chilvers, 2006; Demski et al., 2015; Gregory et al., 2012;

Pidgeon et al., 2014; Webler et al., 2014), though some emphasize a smaller set of these theoretical principles. For example, Demski et al. (2015) conducted an analytic-deliberative workshop to better understand public values when it comes to system-wide energy transitions with explicit attention paid to representation, procedural rules and information used in the

119 process. We identify and characterize components of three decision processes associated with offshore wind project proposals, then relate our findings from our qualitative analysis to the evaluation components from Abelson et al. (2003).

Our investigation of community engagement processes that worked well and those that could be improved focuses on three New England islands at the forefront of offshore wind debates due to their locations near proposed wind farm sites as well as economic and cultural connections to adjacent ocean spaces (e.g., reliance on fishing, sense of place reinforced by aesthetic views).

Due to their proximity to the first offshore wind projects in North America, New England island residents are likely to be among the first positively and/or negatively impacted by this technology.

Three questions drove this work and were also relevant to our community partner, the non-profit

Island Institute. Given the public engagement already occurring in New England on developing offshore wind: 1) What worked well regarding the process of community engagement and its outcomes near proposed offshore wind farms near three New England islands? 2) What were the major challenges with community engagement in these contexts? 3) What insights on community engagement likely apply elsewhere as renewable energy infrastructure proposals become more common? How this industry and other low carbon energy technologies unfold has implications for the rate at which carbon emissions from electricity production are reduced and the timing and extent to which we address climate change.

120 5.2 Methods

Our three pragmatic research questions informed how we collected qualitative data from interviews and relevant documents (e.g., meeting minutes, newspapers, magazines and online news articles), iteratively reviewed and coded the data, compared and contrasted the experiences on three islands, identified common themes, and then related these themes to a theoretical framework, specifically Abelson et al.’s (2013) key components of public participation in deliberation. We identified ways in which our findings resonate with and differ from these components in the analytic-deliberative literature.

5.2.1 Context of study: collaboration with community-based organization

Our project was based on a collaboration between academic social scientists and staff of a non- profit community development organization, Island Institute. This organization has advocated for meaningful public engagement during decision-making processes, including those involving island communities and offshore wind. Using various media, business and community-based strategies, Island Institute has engaged local stakeholders, developers, scientists, engineers, state and federal agency decision-makers and others to learn from each other and consider the trade- offs involved in various development proposals. The Community Energy program staff at Island

Institute has worked with New England coastal and island communities on energy issues since

2008. Our aim with this project was to co-produce knowledge relevant to the communities with which Island Institute works and academic audiences.

We selected three islands based on Island Institute’s long-term engagement with community members, government authorities and wind farm developers involved in the consideration of

121 offshore wind near these particular islands. The proposed wind farms near Block Island,

Martha’s Vineyard and Monhegan Island (see Figure 5.2) are at different stages of project development. The company Deepwater Wind began constructing the Block Island Wind Farm in the summer of 2015. The Vineyard Power Cooperative officially partnered with Offshore MW, a

European wind farm company, in January of 2015. Together, they obtained a lease from the

Bureau of Ocean Energy Management (BOEM) to develop their project in federal waters 12 miles south of Martha’s Vineyard. The University of Maine was not successful in its 2014 bid for funding from the U.S. Department of Energy (DOE) to develop a deep-water floating offshore wind test site near Monhegan Island, but they did secure a smaller grant to continue refining the design of their turbines and they may yet receive a larger DOE grant ($40 million) to deploy and study a full scale prototype (Turkel, 2016).

122

Maine

Vermont

Monhegan Island, Maine New Hampshire Wind resource poten8al Martha’s Vineyard, Massachuse

~~Massachuse~~

~~Block Island, Good Rhode Island Connec8cut Excellent Outstanding~~

~~Figure 5.2. Map of focal islands . Wind data and categorization from NREL (2015).~~

~~5.2.2 Data collection and analysis~~

Island Institute staff conducted unstructured, key informant interviews to collect impressions, opinions and experiences of people closely involved with community engagement in our study sites. These included interviews with town council members, community leaders, government agency employees, leaders of an electricity cooperative and wind farm developers.

We conducted participant observation in that one academic co-author (SK) was hosted by the non-profit organization for 2.5 months to develop a collaborative relationship with Island 123 Institute staff and collect data via informal interviews with them as well as analysis of Island

~~Institute documents and online materials. We also made site visits to the study islands.~~

~~The document analysis involved compiling relevant newspaper articles, reports, meeting minutes and information from websites pertinent to offshore wind and community engagement initiatives.~~

These initiatives were sorted into two categories, namely those that worked well, which research participants associated with legitimacy and positive affect, and those that did not work well, which were associated with expressions of frustration or other negative affect. The academic researcher coded the interview notes and other documents based on qualities associated with stakeholder satisfaction or lack there of, discussed initial themes with Island Institute partners and refined the themes based on their discussions. Finally, these themes characterizing engagement processes that worked well and those that did not work well were compared and contrasted with analytic-deliberative literature on key components of public participation in deliberation.

~~5.3 Results and discussion~~

~~Participants tended to be more satisfied with engagement processes that involved bi-directional and accessible deliberative learning and the provision of custom-tailored community benefits.~~

~~Block Island and Martha’s Vineyard had largely successful community engagement processes resulting in sufficient community buy-in, which contributed to the projects proceeding.~~

Monhegan Island was challenged with a compressed timeline and other initial challenges in building community support. Our interviews, document analysis led us to identify two overarching themes associated with perceptions of legitimate outcomes: accessible, deliberative

124 learning opportunities and community benefits. We then suggested ways to adapt and augment key components of public participation in deliberation to siting renewable energy infrastructure to better incorporate community benefits.

~~5.3.1 Focal island communities and wind farm engagement experiences~~

~~Our island communities differ from those connected by bridges or on the mainland largely based of their relative isolation. We summarize basic characteristics of our three island communities in~~

~~Table 1.~~

~~125~~

Table 5.1. Key differences between New England island study sites and mainland communities relevant to engagement on energy issues. The population and economy characteristics apply to many small towns while energy costs on islands tend to be higher than on the mainland.

Characteristic Description Consequences Year-round Small compared to adjacent Few technical experts Population mainland communities Local leadership positions are often part time or • Block Island: 1,051 volunteer positions • Martha’s Vineyard: 16,535 • Monhegan: 69 (U.S. Census, 2010) Economy Strong dependence on fishing Relatively vulnerable due to low economic and tourism diversification Highly seasonal Year-round residents are likely more available to participate in engagement efforts during low season while seasonal residents and visitors are more likely to engage during the summer Energy Costs Can be higher than mainland, Strong interest in alternatives that could reduce e.g., residential electric rates energy costs, particularly on islands without a grid on Monhegan Island are connection ~$0.70 per kWh and ~$0.15 on the mainland

Richer descriptions of the context for each island’s engagement with offshore wind, including direct quotes from interviewees, are in Appendix Q. Below, we provide a brief overview of engagement processes relevant to the islands we studied.

~~5.3.1.1 Block Island: the ocean state’s offshore wind farm pioneers~~

~~Construction began on the first offshore wind farm in North America in the summer of 2015—a~~

~~30-MW, five-turbine wind farm three miles off the coast of Block Island. A formal state-level marine spatial planning process resulted in the Rhode Island Ocean Special Area Management~~

~~Plan, referred to as SAMP (Nutters and Pinto da Silva, 2012). The SAMP was created and~~

~~126 disseminated before the wind farm was proposed. This meant that information about state waters was already readily available and accessible and had been discussed with key stakeholders~~

~~(Nutters and Pinto da Silva, 2012), including the town council of New Shoreham on Block~~

~~Island, which actively followed and contributed to the SAMP process.~~

The developer and the town council discussed the town’s need for additional technical capacity to make the proposed project more accessible and understandable to residents. The town selected and hired consultants to represent their interests and the developer, Deepwater Wind agreed to reimburse the town for the expense of these consultants (Island Institute, 2012a). Also,

~~Deepwater Wind hired a liaison who had grown up on the island and was well respected by the local community to facilitate community involvement and hold informational meetings.~~

Questions about perceived objectivity (or lack thereof) did not arise in relation to these hires during our analysis. These consultants served the function of a bridging organization between the developers and the island community members. The consultants translated pertinent technical details and locally relevant information to the town council. They shared information with the broader community, fielded questions at community meetings, listened to community concerns and translated these concerns into comments during the formal regulatory processes. The expertise of the consultants provided the town council with greater confidence that community concerns would be better integrated into the wind farm planning processes.

Local stakeholders, government officials and Island Institute staff were convinced that locally- relevant community benefits played an important role in the success of this project. For example, the Block Island wind farm development was done in conjunction with connecting the island to

127 the mainland electricity grid for the first time. The town negotiated to have fiber optic strands included in the underwater electricity cable bundle that now connects Block Island to the mainland grid. Residents and business owners report benefiting from his high speed internet.

~~Deepwater Wind and New Shoreham have also developed a formal Community Benefit~~

Agreement (CBA) in which the wind farm company will pay for improvements to town infrastructure where the cable comes ashore. Further, the project is expected to generate 300 jobs during the construction phase, including opportunities for local mariners and fishermen (Smith et al., 2015).

Block island no longer needs to transport and burn approximately one million gallons of diesel fuel per year to power the island’s generators (Economist, 2015). The island will rely primarily on electricity generated from the wind farm, they will sell excess electricity on particularly windy days and draw from the mainland utility when the wind farm is not operating. The existing diesel system will remain on the island in case of cable failure. There has been some discussion that this system be used occasionally if requested by mainland utilities, in which case they would export some power back onto the cable during heavy load conditions.

~~5.3.1.2 Martha’s Vineyard: moving forward with a cooperative approach~~

~~Vineyard Power grew out of Martha’s Vineyard’s Island Plan, a sustainability strategy that the~~

~~Martha’s Vineyard Commission completed based on input from thousands of island residents in~~

~~2009 to “create the future we want rather than settle for the future we get” (MVC, 2009, p. 1).~~

~~This plan included a recommendation to create a community-owned renewable energy cooperative so islanders could have more autonomy over their energy production and better~~

~~128 ensure community benefits associated with renewable energy development. In 2009, Vineyard~~

Power began recruiting members. People joined for social reasons (e.g., inclusion in the decision making processes in an island-owned, action-oriented group to create a more sustainable energy future for their community) and financial rewards (e.g., ownership and control of local renewable energy projects and stabilized electricity prices once a large-scale renewable energy project is developed) (Nevin, 2010). The cooperative’s community benefits are embedded in the cooperative’s mission: “to produce electricity from local, renewable resources while advocating for and keeping the benefits within our island community” (VPC, 2015).

The cooperative has played an active role in engaging community members in the wind farm decision process. They hosted an interactive offshore wind map viewer on its website to not only inform but also solicit preferences from coop members and other engaged island residents to find a suitable location for the wind farm. This website provided readily available and appropriate information while encouraging participation in sharing local values related to proposed locations.

The website provided information about visual, ecological and human use impacts based on various proposed sites, including data collected from local sources such as island fishermen. The cooperative also hosted a series of community meetings to share wind farm visualizations and solicit feedback (Peckar, 2015a).

In January 2015, BOEM auctioned the rights to lease offshore wind in areas in federal waters south of Martha’s Vineyard. The wind farm developer, Offshore MW, received a 10% discount on their bid price because they had executed a Community Benefit Agreement (CBA) with

~~Vineyard Power. The CBA outlined opportunities to investigate local benefits to the island~~

~~129 including job creation, an operations and maintenance facility, and local equity ownership in the project (VPCOMW, 2015). The size of the wind farm has not yet been confirmed.~~

~~5.3.1.3 Monhegan: confronting deep water and community challenges~~

The tumultuous path of offshore wind in Maine provides insights regarding mutual learning, timing and accessibility of information. In 2009, Maine set ambitious goals to become a national leader in ocean energy (MCP, 2009) and created opportunities for the development of marine renewable energy demonstration projects (MPUC, 2010). Discussions of offshore wind had implications for the island of Monhegan, a remote community 12 miles out to sea with some of the highest energy costs in the nation (MPUC, 2015). In state waters, Maine took initial steps to engage stakeholders in its strategy to expedite the development of the industry by designating three research and demonstration test sites within state waters. State government staff and collaborators hosted a series of public meetings and small and informal discussions along the

Maine coast. They incorporated scientific data and local knowledge into their assessment process by making mutual learning accessible, e.g., traveling to Monhegan where they asked fishermen to rank their fishing activity effort around the island in order to identify a site of least impact for wind turbines.

Efforts to site offshore wind in nearby federal waters underscored the importance of timing and availability of information. The Maine Public Utilities Commission (PUC) began a 16-month process during which they solicited and reviewed bids for and public comments on a long-term power purchase agreement. This extended period of time provided an opportunity to engage stakeholders prior to the announcement of a developer and the location of a site. During this

130 time, the Island Institute worked as a bridging organization to facilitate mutual learning through the Offshore Wind Energy Information Exchange, an outreach and education initiative to inform and engage coastal and marine stakeholders, developers, and decision-makers on the potential for offshore wind energy development in the Gulf of Maine. The initiative included deliberative learning experiences, such as exchange trips to fishing communities as well as a wind farm, the human use mapping project Mapping Working Waters (Island Institute, 2009), information sessions at the annual Fishermen’s Forum in Maine (Island Institute, 2009), and readily available and understandable fact sheets (Island Institute, 2012a). These efforts provided coastal stakeholders and industry representatives with a baseline understanding of community priorities as well as the offshore wind industry, while creating an opportunity for stakeholders to meet each other informally and build relationships.

Maine PUC later announced its selection of an unsolicited proposal from Statoil – a multinational corporation specializing in offshore energy infrastructure – for testing floating turbine technology in federal waters in the state’s Midcoast region. By this time, marine users and other stakeholders in the area had already participated in education and information exchange opportunities, preparing them to more proactively and constructively engage in discussions with the developer and decision-makers (Island Institute, 2015).

The University of Maine entered a federal funding competition with a new scope of activities at the Monhegan test site. Subsequently, the Maine Legislature directed the PUC to reopen the bidding process so that the University of Maine could submit a proposal on an accelerated timeline, and Statoil withdrew its proposal for a project in federal waters. While these

131 developments had statewide implications, this impacted Monhegan by significantly limiting the timeframe in which the community could learn about the change in scope from small-scale portable to large-scale, semi-permanent turbines – a 12 MW pilot project. The PUC opportunity, which prompted many islanders to learn of the change in project scale, was announced during the summer—the island’s busiest time of year.

(METF) to prioritize information that the community needed and facilitate discussion of community benefits associated with the proposed offshore wind project. METF and MAV engaged in weekly phone calls to enhance the flow of information and worked to develop an expectations document to ensure timely project communications. During this time, both parties looked to Block Island for examples of how information was shared and community benefits arranged. MAV also began to host semi-regular open house sessions on the island during which residents and visitors could have more extended discussions about aspects of the project. In late

2015, MAV received additional federal funding ($3.7 mill) to continue refining their floating turbine designs (Turkel, 2015). Based on our interviews, some residents still have concerns about the Monhegan offshore wind project but the developer and community have laid a more solid foundation upon which future communication can take place.

~~132 5.3.2 Bi-directional deliberative learning and community benefits as key to good engagement~~

Our qualitative analysis suggests that much of the myriad considerations for good analytic deliberative processes and outcomes boil down to two key, integrative themes: ensure bidirectional deliberative learning and custom-tailored community benefits. These two overarching themes emerged from our iterative coding process (in which interview notes, attitudes and opinions of various parties involved, engagement materials, meeting minutes and newspaper articles were reviewed, categorized and discussed). Table 5.2 characterizes the two overarching themes of bi-directional learning and community benefits. We discerned four dimensions within the bi-directional learning theme: readily available and appropriate information, trusted messenger, collaboration with bridging organizations and timing of engagement. Reading vertically from the left hand side of the table, it is evident from Table 5.2 that these common themes and associated dimensions played out in various ways across sites.

~~133~~

Table 5.2. Summary of good practices and challenges related to community engagement. For more detail on engagement in three proposed offshore wind farm sites, see site descriptions and Appendix Q. Sites Block Island, RI Martha’s Vineyard, MA Monhegan Island, ME - Town hired consultants to listen, translate and represent community - Vineyard Power Cooperative hosted Readily - Island Institute developed peer-reviewed interests interactive offshore wind map viewer to available & fact sheets to address the questions raised - Developer reimbursed town for inform participants about appropriate during community meetings (Island consultants environmental, human use and visual information Institute, 2012b) - Developer prioritized outreach to impacts community (Island Institute, 2012a) Trusted - Developer hired local liaison to lead - Cooperative founders and members - Leaders in Monhegan Energy Task Force messenger outreach are island residents assumed role of messengers Collaboration - Partnership between local cooperative - Island Institute served as bridging - Consultants helped to bridge town and with bridging and developer provides a bridge to the organization between developer and developer organization community communities - Project preceded by RI Ocean Special - Process to create Martha’s Vineyard - Information Exchange site visits enabled Area Management Plan (SAMP) process, Island Plan and energy coop entailed diverse stakeholders to meet repeatedly which was funded and supported by substantial learning and sharing of and exchange information and federal, state and private entities (Nutters information and values experiences and Pinto da Silva, 2012) - Engagement with fishing industry - Mapping Working Waters project continued after SAMP completed - Coop used online wind map viewer to engaged fishermen to share local - Community meetings from 2009-2012 to solicit resident preferences for farm knowledge and provided opportunity for create and adopt comprehensive energy location them to learn about wind farms (Island plan for Block Island (IEC, 2012) Institute, 2009) Timing of

Bi-directional Deliverative Learning - Formal community engagement from engagement: - SAMP process made information about - University of Maine collected information 2006 to 2010 to create comprehensive, Iterative and state waters readily available before OSW on turbines’ proximity to fishing areas, proactive Island Plan on various multi-year farm was considered (Nutters and Pinto da created and shared visualizations, and sustainability issues Silva, 2012') conducted tourism impacts study

- Timing of engagement around state - Having participated in SAMP process, waters test site activities created - Recruited energy coop members over offshore wind was not a new topic to local challenges from which the community multiple years starting in 2009 leaders when project was proposed organized Monhegan Energy Task Force emerged - Presentations about OSW in both winter and summer to reach year-round and seasonal residents - Embedded in Vineyard Power - Island fishermen were hired to assist with - Provides mainland grid connection Cooperative’s mission and environmental monitoring and site - Reduction in electricity rates organizational structure assessment Deliberation to determine - Ends need to import 1 mill gallons of - Preliminary discussions have included community diesel annually (Economist, 2015) - - Coop members steer siting decision possibility of mainland grid connection, benefits with Fiber optic strands in cable bundle (VPCOMW, 2015) reduced electricity rates, improved flexible models provided to increase internet speed broadband internet for custom - On-island infrastructure improvements - More engagement is needed to more - Community Benefit Agreement tailored benefits - Local jobs provided: mariners and precisely identfy locally appropriate enabled developer to get discount on

~~Provision of Community Benefits fishermen hired to provide security during community benefits lease of ocean space construction~~

~~134 5.3.2.1 Defining bi-directional deliberative learning~~

~~Based on our interviews and document analysis, bi-directional deliberative learning opportunities improved stakeholder engagement in offshore wind project consideration and site development.~~

We use the term bi-directional in reference to mutual learning among developers, government authorities and local community members. Deliberative learning is the exchange of both knowledge and values in a group setting, which is important for developing trust, mutual respect and reaching more satisfying outcomes among those engaged in decision-making processes

(Gregory et al., 2012). Numerous interviewees emphasized the importance of the developers learning about local knowledge, values and priorities. Staff at boundary organizations involved commented on the need to build a shared vocabulary (e.g., megawatt, microgrid) when considering future energy scenarios on each island.

Island Institute staff explained their motivation for their Working Waters participatory mapping exercise as collating different types of knowledge with the goal of sharing facts and values to help address an often unequal power dynamic between project proponents “from away” and local communities. Based on our analysis and relevant literature, wind farm proponents tend to benefit from community engagement strategies in which they learn from the relevant experiences and knowledge of people who could be directly impacted if the proposed development moves forward (see J. Field, 2014).

From our qualitative analysis, we characterized four linked components that we categorize under bi-directional deliberative learning: readily available and accessible information, employment of a trusted messenger/communicator, collaboration with bridging organizations and timing as

~~135 related to iterative learning opportunities over multi-year time frames. These topics arose in numerous interviews and documents as being crucial to the quality of community engagement.~~

~~5.3.2.1.1 Readily available and accessible information~~

Island Institute staff and local government officials in our study sites—echoing the academic literature (Bell et al., 2005)—emphasized how people in adjacent communities needed easy access to information in order to have informed opinions about the proposed wind farms. On the three islands we studied, this information included background on wind farm technology, specifics of a proposed project and how this development could impact individuals and their communities. Island Institute staff and government authorities recognized how skill is needed to translate technical scientific and engineering facts into language that helped lay people learn without being alienated. They stressed the importance of using language accessible to public audiences (e.g., translate megawatts generated into how many average households’ electricity needs will be met in a year, explain what a cable to the mainland means for island residents, explain a power offtake agreement). Island Institute responded to community concerns about a lack of accessible information by creating wind farm fact sheets available in paper form and online (Island Institute, 2012a). Wind farm information in our study sites was published in locally popular newsletters, posted on bulletin boards, paper copies were provided in public places and information was posted online.

~~Island Institute staff compiled local knowledge in their Mapping Working Waters project (Island~~

~~Institute, 2009) because they recognized local knowledge and values need to be translated for wind farm project proponents, marine spatial planners and others working at regional and larger~~

136 scales to better understand the salience, credibility and legitimacy of local perspectives. This type of local knowledge translation, such as fishermen’s expertise on suitable routes to lay the cable (J. Field, 2014) and the location of prime fishing areas to be avoided, is also documented in academic literature as helping to reach legitimate decision outcomes (Failing et al., 2007;

Gregory et al., 2012). The accessibility of information provided during these decision processes was critical given that new information can influence opinions, especially when there are high levels of uncertainty related to a proposed project (Dietz and Stern, 2008) and in situations with widespread misconceptions (Bell et al., 2005; Ottinger and Williams, 2002).

~~5.3.2.1.2 Trusted messenger~~

Our interview data showed that Block Island wind farm developers recognized the importance of hiring a trusted liaison from the local community to help facilitate the community engagement process. Communication between community members and project proponents was an issue on

~~Monhegan Island, for various reasons including the compressed time frame to submit a federal grant proposal and, potentially, because the developer had no local, Monhegan-based staff.~~

Consequently, more effort has been invested in relationship building, particularly between the developer and a community energy group, the Monhegan Energy Task Force as the developer prepares to apply for additional funding.

Our interpretation of the central role of trusted communicators aligns with numerous studies that have documented how the messenger matters of information may matter more than the information delivered (Cialdini and Goldstein, 2004; Kahan, 2010; 2012; Wynne, 1992). Studies have shown that if a technology and its costs and benefits are not appropriately translated or

137 people distrust the source of the information, stakeholders may feel alienated or disengage from the decision process (Wynne, 1992; 1989), and potentially become entrenched in their opinion regardless of new information that arises (Kahan, 2010). Information alone has a limited influence on opinions (Kahan et al., 2012). People tend to “endorse whichever position reinforces their connection to others with whom they share important commitments” (Kahan,

2010, p. 297). Arguably more important than technical information, the social context in which information is shared and the person presenting it—the messenger—can exert substantial influence on attitudes, opinions and behavior (Cialdini and Goldstein, 2004; Kahan, 2010). This encompasses the personalities, communication styles and values of people sharing information and facilitating community meetings and dialogues.

~~5.3.2.1.3 Bridging organizations~~

Island Institute as a bridging organization spearheaded participatory mapping of fishing effort to inform marine spatial planning (Island Institute, 2009). Part of the rationale for this project was to shift local stakeholders from playing the role of recipients of information to producers of information that developers and government officials could understand, respect and use. Tobias

~~(2009) documents how boundary organizations can help provide such potentially empowering experiences for local stakeholders.~~

The experiences of people involved in our offshore wind farm study sites reinforce the critical role that boundary organizations can play in supporting community engagement. Echoing Cash et al. (2006), boundary organizations assisted in the co-production and sharing of knowledge for

~~138 decision-making in our study sites. Boundary or bridging organizations can be defined with the following characteristics (Cash et al., 2003):~~

• Accountability to both sides of a boundary, e.g., local communities and project proponents • Use of “boundary objects,” e.g., maps reports, and forecasts, which actors on different sides of a boundary co-produce • Participation across the boundary involving - Convening - Translation - Coordination of complementary expertise - Mediation

Island Institute, SeaPlan, Gulf of Maine Research Institute and NOAA’s Sea Grant program are examples of bridging organizations that played important roles in relation to the island communities that we studied. Interviewees characterized them as more objective third parties

(i.e., more objective than the developers). These organizations helped run community engagement and public outreach processes related to marine spatial planning and offshore wind farm siting, but did not push for specific outcomes. On Block Island and Martha’s Vineyard, our interviews and document analysis showed that project proponents and local government retained organizations and people with excellent communication and facilitation skills who the community already trusted. It is likely that part of the success of using these relatively neutral people who served as communication bridges is that stakeholders are more likely to be open to learning new information if the values of the messenger and/or bridging organization resonate with them (Kahan, 2010).

~~139 5.3.2.1.4 Timing: substantial iterative public engagement before site selection~~

Iteration emerged as a requisite characteristic of the community engagement processes characterized by minimal participant frustration. These iterative learning opportunities unfolded over multiple years. They involved joint fact-finding, such as Rhode Island’s Special Area

~~Management Plan process, and values clarification, such as the prioritization of sustainability issues and potential solutions in the Martha’s Vineyard Island Plan.~~

Timing was problematic on Monhegan Island. From our interviews, we surmise that in all three sites developers were often reluctant to share uncertain details, such as the specific location of a site, before they were confirmed. During an early stage of the project, developers on Monhegan

Island tended to share only incomplete information when they engaged in community meetings, which frustrated local stakeholders, some of whom perceived the developer as being dishonest by withholding information. The uncertainty of the impacts also frustrated stakeholders.

The frustration that select interviewees expressed suggests that some public mistrust, skepticism and opposition to the Monhegan renewable energy proposals may have been (or could be) reduced with more frequent, meaningful and timely opportunities for locals to voice their concerns in decision-making (Bell et al., 2005; Gregory et al., 2012). Literature on planning processes and environmental management stresses the importance of engaging communities early and often (Dietz and Stern, 2008; Gregory et al., 2012) yet, as our island examples show, this can be challenging due to uncertainties inherent in early stages of project development. It became apparent from our research that wind farm developers often spend years collecting the requisite information to comply with regulatory requirements and determine optimal sites.

~~140~~

Upstream research engagement can help navigate uncertainties associated with a new technology and the impacts it may have. Scholars are beginning to study upstream deliberation regarding offshore renewable energy (Wiersma, 2016; Wiersma and Devine-Wright, 2014). When conducting upstream research, scientists, government authorities, bridging organizations and/or developers can discuss a new technology with citizen groups before any choices are made regarding if and where the technology may be used. Upstream research can help scientists and developers to “open innovation processes at an early stage to listen, respond and value public knowledge and concerns related to risks and ethical dilemmas” (Wilsdon and Willis, 2004, p.

~~28). This type of research can help answer people’s questions, including “Why this technology?~~

Why not another? Who needs it? Who is controlling it? Who benefits from it? Can they be trusted? What will it mean for me and my family? What are the outcomes that this technology seeks to generate? Could we get there in another, more sustainable and cost-effective way?”

~~(Wilsdon and Willis, 2004, p. 28).~~

We recommend that when state, tribal and federal agencies initiate ocean planning, they also facilitate upstream research as pertains to potential new uses of ocean space that may not yet be pressing issues. Ocean planning involves coordinating regional planning for current and future ocean industry, conservation and recreation. Before areas are designated for specific ocean uses, such as offshore renewable energy development, ocean planning initiatives have provided opportunities for data collection, dialogue on various uses and values and sharing of information.

~~More of this kind of early engagement could help stakeholders learn about technologies and how they could be managed without triggering place-protective opposition. Such opposition can stem~~

~~141 from perceived threats to specific places that may be important to people’s sense of identity and to which they may have other strong attachments (Devine-Wright, 2009).~~

In addition to being included in ocean planning processes, BOEM also has the potential to facilitate upstream research as the agency interacts with state, tribal and local governments through task force meetings on specific offshore resource issues. This helps in providing transparency regarding issues at different levels of government and provides opportunities for stakeholders to learn and ask questions about areas of federal waters or specific projects. BOEM has the authority to collect and share data on and then define boundaries of offshore ocean areas that are available via leases to wind farm developers (Firestone et al., 2015). Through BOEM’s task force meetings, information is directed to the specific set of stakeholders that an offshore renewable energy project may affect. This type of early engagement with stakeholders is critical in any ocean development project.

~~Our interviewees emphasized how early engagement dispelled community member’s fears of finding out too late to become meaningfully involved in decision processes on Martha’a~~

Vineyard and Block Island. On Martha’s Vineyard, the steps of the process and the timeline for making various decisions related to island sustainability in general and later offshore wind enabled stakeholders to understand how and when to engage in the process. Boundary organizations, developers, and government agency staff recognized time and resource challenges around iterative and potentially multi-year stakeholder involvement in decision processes. Our analysis showed that building trust among proponents, the selected ‘messengers’ and communities takes time as does allowing for new information and questions to arise. Based on

~~142 the literature and our qualitative analysis, timely deliberation on identifying and procuring community benefits can also build trust.~~

~~5.3.2.2 Provision of community benefits~~

Island Institute staff, community leaders and local government officials thought that explicit inclusion of community benefits was key to successful engagement processes on Block Island and Martha’s Vineyard. Engagement efforts in Monhegan did not include substantial discussion on this topic prior to 2016.

By community benefits, we mean additional and distinct funds or investments that the developer provides to communities, often near project sites (B. J. A. Walker et al., 2014). Benefits associated with the generation of renewable electricity, such as carbon reduction, are diffuse and tend to accrue at a global scale while several environmental, economic and landscape impacts are concentrated and local. Providing community benefits above and beyond tax revenues can play an important role in managing renewable energy scale-related distributional conflicts (Wolsink,

~~2007; Zografos and Martínez-Alier, 2009).~~

Whereas the term ‘community benefits’ has been used broadly, the experiences of those engaged with our study sites suggest a need for a more nuanced theorization of this term. That is, whereas the term itself could be viewed from a utilitarian perspective as simply providing net benefit to the majority, our study sites demonstrate that such a narrowly utilitarian approach does not sufficiently capture strongly held community concerns of fairness. Whereas one might think that a community benefits from a project if the majority receives a net benefit, and the community-

143 scale aggregate is a net benefit, our data suggest that these are not sufficient criteria. Individuals expressed concern that specifically impacted groups may require compensation, i.e., some island leaders and boundary organization staff expressed how compensation should be considered for fishermen who would lose fishing grounds. These individuals were not among those who would be most directly impacted by an offshore wind farm. Accordingly, we seek to make explicit that broadly acceptable community benefits are benefits to individuals and groups as seems fair and appropriate from a community perspective.

This qualifier adds a broader relational perspective that integrates not only consequences but also principles and notions of fairness at scales coarser than an individual. From this perspective, individuals may oppose a project even if they might personally gain from it (e.g., a local barge operator may get numerous contracts from an offshore wind project), if they seem unfair at a community level, accounting for the existing and historic relationships and the prevailing values of a place (e.g., the wind farm siting process may not be sensitive to the preferences of local lobstermen).

Community benefits can help balance the provision of private and public benefits associated with an offshore wind farm. Some perceive offshore wind development as privatizing the ocean, which, historically, has been a public space for fishing, recreating and other activities (Devine-

Wright and Howes, 2010; Firestone et al., 2009; Pomeroy et al., 2014). The federal management agency overseeing the development of offshore wind, BOEM, has public good-oriented goals, but they use market-based tools to achieve these (e.g., auctions involving private developers).

~~Part of BOEM’s mission is to “promote energy independence, environmental protection and~~

144 economic development” via delineating and auctioning areas of the ocean for different purposes, including offshore wind farms (BOEM, 2015). We suspect that BOEM’s general public good- oriented goals are less salient to residents of communities adjacent to wind farm sites compared to local concerns, such as displacement of fishermen from fishing grounds, but we did not measure this (Island Institute, 2012b). In order to shift perception of benefit from the large scale and general to the local and specific, developers may provide community benefits for various reasons, such as to help earn the public’s trust and create a sense of fairness associated with the project (Aitken, 2010; Cowell et al., 2011; Rudolph et al., 2015). However, as noted in European case studies, the formation and provision of community benefits can erode or build trust and perceptions of fairness (Aitken, 2010). Community benefits literature and our research demonstrate how establishing trust and perceptions of fairness rest on both the process of coming up with appropriate benefits as well as the models and mechanisms used to deliver the benefits.

~~5.3.2.2.1 Deliberation to determine community benefits~~

Relevant literature and our island-focused research point to the importance of collaboration among developers, communities and government agencies to identify and provide community benefits rather than only respond to government mandates about benefits (Aitken, 2010; Rudolph et al., 2015). Community benefits are required by law in some contexts and voluntary in others.

For example, land-based wind developers in Maine must pay host communities according to the number of installed turbines (Maine State Legislature, 2010) but offshore wind developers are not required by law to provide community benefits in the UK (Aitken, 2010).

145 Our research and relevant literature supports how early discussions among government authorities, developers and communities are needed to arrive at acceptable definitions and understandings of communities, benefits, impacts and how they relate to each other (see Figure

5.3). We have thus far used the term community in reference to residents of particular islands, but communities can be based on location (e.g., a town), interests (e.g. recreational boaters), groups who are adversely impacted (e.g., commercial fishermen), organizations (e.g., an energy cooperative) and/or other shared characteristics. Benefits can be understood as sharing economic gains associated with tapping into a public natural resource (i.e., wind), recognition of hosts

(e.g., developer seeks to be a good neighbor, communities receive benefits for hosting substation infrastructure), increasing local support (e.g., community groups or energy cooperatives who receive benefits commit to supporting wind farm), accounting for impact (e.g., recognition of local negative impacts), compensation for agreed upon and specific losses (e.g., funds to improve habitat for birds at high risk of collision with turbines). Impacts can be perceived as positive

(e.g., provision of jobs and carbon neutral electricity) and/or negative (e.g., bird mortalities, decreased visual amenities). Rudolph et al. (2015) developed a framework to achieve the legitimate provision of community benefits via a set of interactions among communities, benefits and impacts (Rudolph et al., 2015). Community engagement processes on two of the islands we studied had substantial community support (Martha’s Vineyard and Block Island) and covered the topics in this framework when they developed community benefits. Interviews documented that wind farm developers for the Monhegan project have come to recognize the role of community benefits in the other islands’ development processes and are working towards discussion about what such benefits could be for Monhegan.

~~146 Government Communi;es Developers Authori;es Stakeholders~~

Who should benefit? Beneficiary communi2es can be defined by - Loca2ons: town, island - Interests/prac2ces: fishermen, sailors - Groups adversely impacted: fishermen - Organiza2ons: energy coopera5ves, conserva5on groups - Other aJributes: demographic characteris5cs

Why & how to What provide beneﬁts? are the impacts? - Share economic - Environmental gains associated with - Social using public resource - Economic - Recognize hosts How are impacts - Account for impact - Compensate for perceived? speciﬁc losses - Posi2vely - Other - Nega2vely

~~To collabora2vely develop Appropriate Community Beneﬁts~~

Figure 5.3. A robust approach to developing community benefits. This requires reaching a common understanding of impacts, communities, fair and appropriate benefits, and their interactions among developers, communities and government authorities. Italics denote examples. Adapted from Rudolph et al. (2015).

~~147 5.3.2.2.2 Flexible models for custom tailored benefits~~

Community benefits took different forms in our three study sites. They can be integrated into various stages of a project, such as the planning, permitting, mitigation, operational and decommissioning stages. We add to Rudolph et al.’s (2015) overview of common offshore wind community benefit models and mechanisms:

• Community funds (most common) • Other and pre-existing funds • Community ownership • Equal distribution of revenues • Direct investment and project funding (e.g., paying for infrastructure improvements) • Jobs, apprenticeships and studentships • Educational programs • Electricity discounts • Community benefit agreements • Indirect benefits from the supply chain • Indirect benefits via tourist facilities

It may be instructive for communities, government authorities and developers to look to Europe when considering appropriate community benefits. In Denmark and regions of Germany, community benefits are often based on cooperative models in which members own the business and all profits after taxes are given back to members (Breukers and Wolsink, 2007). In the UK, energy developers annually pay into a fund proportional to the megawatts (MW) of installed capacity for community organizations to spend on local initiatives (Cowell et al., 2011). For more detailed descriptions of different types of community benefits, see Rudolph et al. (2015).

Community benefits have the potential to enhance or degrade relationships between developers, government authorities and local communities; they can be perceived as broadly beneficial or a bribe that displaces civic duty (Sandel, 2012; B. J. A. Walker et al., 2014). Co-creating community benefits so they are perceived as fair and appropriate from a community perspective

148 may reduce the perception among stakeholders of benefits as bribes. Establishing locally- appropriate community benefits involves clearly identifying their scale, role and purpose in order to reduce this potential negative perception (Cowell et al., 2011). This process can also improve clarity and diminish uncertainty about what will be provided so developers can discuss them earlier in the planning stages. Rudolph et al. (2015) recommend that developers and authorities negotiate with communities about various benefit models during early stages of wind farm planning, ideally before submitting planning applications.

~~5.3.2.3 Relevance to components of public participation in deliberation~~

We conducted our qualitative analysis before reviewing principles for public participation in deliberation. Many of the concepts that emerged from our analysis associated with successful and/or frustrating parts of engagement processes reinforce principles from Abelson et al. (2003).

The principles from Abelson et al. (2003) that arose more than once in our qualitative analysis are outlined in blue in Figure 5.4. We augment these principles with consideration of community benefits in deliberative processes that may result in an imposition of one party’s interests on a community (e.g., wind farm developers interests imposing on adjacent community member’s interests). It is likely that Abelson et al, (2003) did not attend to community benefits because the topic of their review was health policy and the presumed community benefit was improved health. Explicit attention to community benefits, as depicted in orange boxes in Figure 5.4, could apply broadly to community engagement with various types of infrastructure and technology, not just to a developer building a wind farm.

~~149 Legi)mate and fair selec)on process~~

~~Access to par)cipa)on in decision process Demographic~~

Representa)ve sample Geographic Legi)mate Representa3on Clarity in Poli)cal Fair selec)on process Par)cipant Responsive Key selec)on vs. self- Reasonable selec)on characteris)cs Major features, e.g., Public input agenda seEng sought Minor features, e.g., Ample )me for order of who speaks discussion Key components Procedural rules Organiza)onal Who is listening of public to public? par3cipa3on in level of public par)cipa)on delibera3on Who responds and input to public? Par)cipants have opportunity to challenge process, informa)on presented and experts Informa3on involved used in process Appropriate selec)on Characteris)cs Outcomes and and presenta)on decisions arising Public input Accessible, readable, from process: incorporated Fair selec)on Legi3macy and into final diges)ble accountability for decision Sufficient )me to Who chooses outcomes consider, discuss informa)on Decisions & and challenge public input informa)on Community Who chooses experts into them are provided Benefits communicated that provide informa)on to the public

Ci)zens are more informed about issues Iden)fy public input that was incorporated Shared Iden)fy and Decision-making authority responds to understanding provide custom- public input Answer why public input across tailored beneﬁts was incorporated or not deemed fair and BeSer or diﬀerent decision communi)es, appropriate from authori)es & a community Broad understanding and developers perspec)ve. acceptance of decision

~~Iden)fy Who should Why and how to impacts beneﬁt provide beneﬁts~~

Figure 5.4. Design and evaluation principles for public participation processes with community benefit outcomes. Blue outlines denote topics from Abelson et al. (2003) that arose in multiple interviews and our document analysis despite how we did not provide specific prompts for these topics. Orange denotes attributes of community benefits that were perceived as crucial to the success of the wind farm decision processes that we studied. We recognize the importance of topics in black outlines from Abelson et al. (2003), even though they were not common topics in our interviews or document analysis.

~~150~~

~~5.4 Conclusion~~

Proposals for renewable energy infrastructure are poised to rapidly proliferate, particularly if countries follow through with carbon reduction commitments. The ways in which humanity approaches, manages and responds to inevitable controversy over these technologies impacts the pace and efficacy of addressing climate change and transitioning to low carbon energy sources

(Roberts et al., 2013). Based on results from the islands we studied and literature synthesis, we see the critical importance of developers and decision makers engaging local communities to address concerns about project impacts and benefits to achieve legitimate decision outcomes.

~~Communities may legitimately reject particular renewable energy technologies.~~

Furthermore, we augment established principles for public participation in deliberation that focus on process with an explicit inclusion of a particular outcome. Specifically, if the project is considered worthy of moving forward, we recommend outcomes of community benefits deemed fair and appropriate by communities that incorporates viewpoints from government authorities and developers.

Deliberative analytical decision processes involving extensive stakeholder engagement can be resource and time intensive, but this initial investment can result in lower long-term costs with potentially fewer delays, it may reduce the risk of litigation costs (Irvin and Stansbury, 2004;

~~Randolph and Bauer, 1999) and we suggest it may result in better long-term relationships among those involved. Based on what we learned from the experiences of Block Island, Martha’s~~

~~Vineyard and Monhegan Island, building a foundation of both knowledge and trust is crucial for~~

151 the success of an offshore wind farm and likely other renewable energy technologies. Making deliberative learning accessible and providing clear community benefits can help ensure that 1) the decision-making processes around these projects are inclusive, effective and perceived as fair; 2) local, scientific and political knowledge is considered; and 3) projects that are considered appropriate after an analytic-deliberative process are properly sited.

~~152 Chapter 6: Conclusion~~

The purpose of my dissertation was to shed light on controversies and potential solutions at the confluence of climate change and the biodiversity crisis in a manner that addresses human psychology, including fundamental desires for connection to others, both human and non-human.

~~I am convinced that such insights can help combat the currently bleak data trends and unfolding disasters resulting from climate change and dramatic reductions in biological diversity and abundance.~~

Scientists and economists have attempted to cut through decades of political debates on the validity of climate change by repeatedly calling for “vigorous efforts to develop low-carbon technologies” (2013, p. 326). Countries implemented this recommendation to a degree, but a vast gulf separates the planet’s current trajectory from the atmospheric conditions we must achieve to stabilize the climate. For example, non-hydropower renewable energy sources are growing faster than any other source for new generation capacity globally. Yet, they comprised only 5% of the total world electricity generation as of 2012 while coal generated 40% (DOE EIA, 2016).

Navigating this gulf entails re-considering societal priorities and the technologies we use. In democratic societies, partial solutions to climate change and the biodiversity crisis require public support, which led me to investigate perceptions of impacts and benefits associated with a technology—offshore wind farms—that could contribute towards reducing greenhouse gas emissions.

~~153 6.1 Realization of renewable energy research goals and research implications~~

This dissertation applies and integrates social studies of risk, ecosystem services, environmental and relational values and theory on analytic-deliberative processes to barriers to scaling up renewable energy. This integration addresses facets of public opposition to renewable energy development based on concerns about social, financial and environmental consequences, value orientations and flawed engagement practices.

Chapter 2 provides preliminary evidence that an important set of insights from risk perception may apply to how people understand environmental impacts. Specifically, I showed that components of the psychometric risk paradigm extend beyond its traditional domain of environmental health concerns (e.g., carcinogens) to indirect impacts to people via concerns about changes to ecosystem services. These findings would suggest that developers and government authorities might anticipate stakeholder concerns more effectively by attending to the characteristics of risks. For example, risks perceived to be uncontrollable, and those that invoke dread—e.g., those associated with fatalities of animals—are more likely to induce higher levels of concern. This preliminary evidence that perceived risk research also applies to ES gives environmental researchers a new set of methods and conceptual tools for understanding the types of environmental impacts that will likely loom large in the public's mind, and which impacts will largely escape notice.

Whereas Chapter 2 focused on risk perceptions, Chapter 3 demonstrated that many people are willing to pay to mitigate the harms of offshore wind farms—and are willing to pay even more for the wind farms to have net positive ecological effects via enhanced habitat for underwater

154 species. The implications of this high latent demand for biodiversity-friendly offshore wind farms is that it suggests willingness on the part of utility payers to fund renewable energy that is not only clean climate-wise, but also ecologically beneficial. This public support could help transform the energy landscape. The strongest preferences and consequently highest willingness to pay amounts were for wind farms with contributions to species richness and abundance that are net positive. People favor the regenerative options and are willing to pay for them. An implication of Chapter 3 is that renewable energy proponents may increase support for their proposed projects by going beyond mitigating risks to biodiversity to making the infrastructure ecologically regenerative.

~~An important caveat for the experiment in Chapter 2 is that the sample that I surveyed~~

~~(Northeastern US) is a relatively environmentally oriented and wealthy part of the world.~~

~~Nonetheless, the choice experiment models point to a surprisingly large willingness to pay for ecologically regenerative offshore wind farms.~~

Recognizing that not all values are monetary, Chapter 4 provided some empirical evidence worthy of further investigation on the emerging concept of relational values. Specifically, I showed that these relational values—values linking people and ecosystems via tangible and intangible relationships as well as the principles, virtues and notions of a good life that may accompany these—may be both strong and widely held. Furthermore, my results indicate that people respond to relational value statements differently than how they respond to New

~~Ecological Paradigm statements, the latter being a common way to assess ecological worldviews.~~

~~I found preliminary evidence that this novel relational construct is predictive of attitudes and~~

155 preferences toward wind farms. Conservation scientists and practitioners may have been missing this important relational dimension of attitudes and values about the environment. Relational values need more testing, but they may be sufficiently cohesive and discrete to be an important construct for understanding environmental values and motivations for pro-environmental behavior.

In Chapter 5, my qualitative analysis demonstrated that, amongst the litany of criteria in the literature, good public engagement in three island communities boiled down to two key themes: enabling bidirectional deliberative learning and providing community benefit. That is, the smoother decision processes included public engagement in dialogue in which participants, including experts and local stakeholders, learned from each other while reconciling technical expertise with citizen values. Outcomes included the provision of community benefits that have important relational dimensions in that these benefits should be collaboratively negotiated. The resulting benefits ought to be perceived broadly as fair and appropriate from a community perspective. Attending to these two key themes may improve the quality of the interactions among communities, government authorities and developers when deciding if and where to site renewable energy infrastructure.

~~6.2 Limitations~~

As an initial foray into several rapidly expanding areas of research, my dissertation of course has several notable limitations. First among these for Chapter 2 is the relative small size of my sample. My exploratory in-depth mixed methods were tested on a small sample (n = 27). As always, a larger sample size would make the statistical analysis and the corresponding findings

156 more robust. Also, in contrast to health risk studies comparing expert to lay people, I had no independent measure of the magnitude of risks to ES. Instead, my scoring system was based on interpreting my interview data and academic publications on this topic. Accordingly, it is possible that some of what appeared to be a strong risk signal may actually stem from the fact that some risks (like those to marine mammals and birds) are actually more likely to be large in magnitude.

~~A potential limitation of Chapter 3 is that my survey respondents (Mechanical Turk workers) are not a fully representative sample of residents of the geography that I targeted (coastal New~~

England states). As I mention in Chapter 3, the demographic characteristics of Mechanical Turk workers differed from census data. I found, however, no evidence that demographic characteristics influenced choices in the experiment. Intuitively and without empirical evidence,

I suspect that these online workers may be more accepting of novel technologies because they have chosen to work within a relatively new online system. Although I did not find obvious biases in how people with different demographic characteristics responded, I would not completely rule out that their relative youth and higher levels of education may have made them more likely to support environmentally friendly renewable energy. The expense of using an online panel that is more closely representative of a targeted population may be worthwhile for future studies on this topic.

~~Another limitation in Chapter 3 was my inclusion of marine habitat impacts in the design of my choice experiment while excluding impacts to birds, which was a highly prominent concern in~~

~~157 Chapter 2. Future research could include wind farm design features with different levels of positive and negative impact on avian life.~~

~~My estimates of willingness to pay (WTP) for artificial reef are higher than those found by~~

~~Börger et al. (2015) who used an additional annual tax to measure WTP and sampled from a population in England potentially impacted by the development of an offshore wind farm.~~

Although I used monthly in the preamble to the survey and in the description of each choice, respondents may not have reflected on how much this tax would cost annually, which may partially explain my high WTP estimates. Future research could evaluate sensitivity to monthly as compared to annual taxes in choice experiments related to offshore wind.

Lastly, I used vivid graphics to convey wind farm characteristics in the experiment. I did not, however, assess if the high willingness to pay amounts could be partially attributed to the bright images used to represent different qualities of artificial reef habitat. It is possible that respondents may have made quick, intuitive decisions to choose the most colorful option without reflecting much on the hypothetical cost incurred from this choice.

While it provides promising findings, Chapter 4 suffers from the normal limitations of a preliminary exploration. Specifically, the six relational value statements each address a potentially separate aspect of values about relationships (e.g., kinship with nature as distinct from responsibility for impacts to others). While these six statements showed consistency as a set (an interesting and somewhat unexpected finding in itself), some researchers will be interested in the sub-component values separately. In the exploratory analysis described in Chapter 4, we did not

158 test whether each of these relational value statements has internal validity. Accordingly, we do not know if slightly re-wording each type of relational value would change how people respond to it.

As is the case with most site-based research, the results of Chapter 5 may be limited in their generalizability across different types of renewable energy infrastructure in different regions of the world. We grounded our assessment in more generalizable academic literature when we evaluated if and how theoretical ideas about analytic-deliberative processes played out in three sites. Given the limited number of sites, our results about bi-directional deliberative learning and community benefits may not strongly resonate in other places. More research is needed to test the generalizability of these qualitative results while recognizing the extensive literatures demonstrating that both procedural and distributive justice matter.

~~6.3 Future research directions~~

Much research remains to better understand the complexity of public support and opposition to sustainable energy transitions. I see a need for more human-centered energy-related research methods (e.g., surveys, interviews, focus groups) to reveal additional underlying factors motivating or hindering the adoption of offshore wind infrastructure and other renewable energy technologies. Such research could also assess why, how and in what contexts attitudes and behaviors towards energy technologies change. Pre and post surveys could be used to investigate energy-related attitudes and behavior changes over time, which could help research and consequently practitioners anticipate future attitude and behavior changes.

159 My research is one of a small handful of academic studies that have focused on community engagement with the nascent offshore wind industry in the US. More extensive and longer-term research that involved additional interviews and becoming more embedded in decision processes relevant to marine planning could lead to additional insights.

Building on my wind farm choice experiment results in Chapter 3, additional research could be done in collaboration with a renewable energy business to assess real willingness to pay, not just hypothetical willingness to pay for renewable energy infrastructure that has regenerative design features.

~~6.4 Towards ecologically and socially sustainable energy~~

The climate negotiations at COP16 established higher emissions reduction targets and more accountability via emissions reporting requirements than preceding international climate change agreements (UN, 2015). These ambitious targets may tempt policy-makers to streamline public engagement processes to deploy the technologies faster. Such streamlining could be counter- productive, potentially increasing the rates of lawsuits and developers losing their social license to operate. My research and other studies reinforce how we need well planned analytic deliberative processes (Devine-Wright et al., 2011). More broadly, echoing Stehr (2016), my research points towards confronting climate change as an opportunity for more democracy, not less.

I identified large, latent support for ecologically regenerative renewable energy technology and a strong suggestion that relational values could help propel a sea-change in actions as well as social practice around reconfiguring our energy systems. Securing sustainable energy may prove

~~160 an illusive goal, but this research has helped bring some social and ecological facets of this goal into sharper focus while providing a foundation upon which future research can be launched.~~

161 References Abelson, J., Forest, P.-G., Eyles, J., Smith, P., Martin, E., Gauvin, F.-P., 2003. Deliberations about deliberative methods: issues in the design and evaluation of public participation processes. Social Science & Medicine 57, 239–251. doi:10.1016/S0277-9536(02)00343-X Abson, D.J., Wehrden, Von, H., Baumgärtner, S., Fischer, J., Hanspach, J., Härdtle, W., Heinrichs, H., Klein, A.M., Lang, D.J., Martens, P., Walmsley, D., 2014. Ecosystem services as a boundary object for sustainability. Ecological Economics 103, 29–37. doi:10.1016/j.ecolecon.2014.04.012 Acheson, J.M., 2003. Capturing the Commons. University Press of New England. Adamowicz, V., Naidoo, R., 2016. Discussion about “cheap talk.” Adams, W.M., 2014. The value of valuing nature. Science 346, 549–551. doi:10.1126/science.1255997 Aitken, M., 2010. Wind power and community benefits: Challenges and opportunities. Energy Policy 38, 6066–6075. doi:10.1016/j.enpol.2010.05.062 Allum, N., Sturgis, P., Tabourazi, D., Brunton-Smith, I., 2008. Science knowledge and attitudes across cultures: a meta-analysis. Public Understanding of Science 17, 35–54. doi:10.1177/0963662506070159 Angelou, N., Elizondo Azuela, G., Banerjee, S.G., Bhatia, M., Bushueva, I., Inon, J.G., Jaques Goldenberg, I., Portale, E., Sarkar, A., 2013. Sustainable energy for all (No. 77889). World Bank, Washington, DC. Ansolabehere, S., Konisky, D.M., 2014. Cheap and Clean. MIT Press. Antunes, P., Kallis, G., Videira, N., Santos, R., 2009. Participation and evaluation for sustainable river basin governance. Ecological Economics 68, 931–939. doi:10.1016/j.ecolecon.2008.12.004 Armsworth, P.R., Chan, K.M.A., Daily, G.C., Ehrlich, P.R., Kremen, C., Ricketts, T.H., Sanjayan, M.A., 2007. Ecosystem‐service science and the way forward for conservation. Conservation Biology 21, 1383–1384. doi:10.1111/j.1523-1739.2007.00821.x Arnett, E.B., Brown, W.K., Erickson, W.P., Fiedler, J.K., Hamilton, B.L., Henry, T.H., Jain, A., Johnson, G.D., Kerns, J., Koford, R.R., Nicholson, C.P., O'Connell, T.J., Piorkowski, M.D., Tankersley, R.D., JR., 2008. Patterns of bat fatalities at wind energy facilities in North America. Journal of Wildlife Management 72, 61–78. doi:10.2193/2007-221 Auster, P.J., Langton, R.W., 1999. The effects of fishing on fish habitat. American Fisheries Society 1–45. Bagheri, H.C., Del Amo, B., 2016. The past and future of renewable energy deployment in light of the COP21 agreement: the cases of Germany and the USA. SSRN Journal 1–13. Baine, M., 2001. Artificial reefs: a review of their design, application, management and performance. Ocean and Coastal Management 44, 241–259. doi:10.1016/S0964-5691(01)00048-5 Barnosky, A.D., Brown, J.H., Daily, G.C., Dirzo, R., Ehrlich, A.H., Ehrlich, P.R., Eronen, J.T., Fortelius, M., Hadly, E.A., Leopold, E.B., Mooney, H.A., Myers, J.P., Naylor, R.L., Palumbi, S., Stenseth, N.C., Wake, M.H., 2014. Introducing the scientific consensus on maintaining humanity's life support systems in the 21st century: information for policy makers. The Anthropocene Review 1–32. doi:10.1177/2053019613516290 Barry, J., Ellis, G., Robinson, C., 2008. Cool rationalities and hot air: a rhetorical approach to understanding debates on renewable energy. Global Environmental Politics 8, 67–98. doi:10.1162/glep.2008.8.2.67 Barry, M., Chapman, R., 2009. Distributed small-scale wind in New Zealand: Advantages, barriers

162 and policy support instruments. Energy Policy 37, 3358–3369. doi:10.1016/j.enpol.2009.01.006 Bell, D., Gray, T., Haggett, C., 2005. The “social gap” in wind farm siting decisions: explanations and policy responses. Environmental Politics 14, 460–477. doi:10.1080/09644010500175833 Bell, D., Gray, T., Haggett, C., Swaffield, J., 2013. Re-visiting the “social gap”: public opinion and relations of power in the local politics of wind energy. Environmental Politics 22, 115–135. doi:10.1080/09644016.2013.755793 Bennett, N.J., 2016. Using perceptions as evidence to improve conservation and environmental management. Conservation Biology 30, 582–592. doi:10.1111/cobi.12681 Berbés-Blázquez, M., González, J.A., Pascual, U., 2016. Towards an ecosystem services approach that addresses social power relations. Current Opinion in Environmental Sustainability 19, 134–143. doi:10.1016/j.cosust.2016.02.003 Bergström, L., Kautsky, L., Malm, T., Rosenberg, R., Wahlberg, M., Astri, Wilhelmsson, D., 2014. Effects of offshore wind farms on marine wildlife—a generalized impact assessment. Environ. Res. Lett. 9, 034012. doi:10.1088/1748-9326/9/3/034012 Bishop, I.D., Miller, D.R., 2007. Visual assessment of off-shore wind turbines: The influence of distance, contrast, movement and social variables. Renewable Energy 32, 814–831. doi:10.1016/j.renene.2006.03.009 Blackmore, E., Underhill, R., McQuilkin, J., Leach, R., 2013. Common cause for nature. Public Interest Research Centre, Machynlleth, Wales. Blackstock, K.L., Kelly, G.J., Horsey, B.L., 2007. Developing and applying a framework to evaluate participatory research for sustainability. Ecological Economics 60, 726–742. doi:10.1016/j.ecolecon.2006.05.014 Boehlert, G.W., Gill, A.B., 2010. Environmental and ecological effects of ocean renewable energy development: a current synthesis. Oceanography 23, 68–81. BOEM, 2015. About BOEM [WWW Document]. httpwww.boem.govAbout-BOEM. URL http://www.boem.gov/About-BOEM/ (accessed 11.30.15). Bohnsack, J.A., Sutherland, D.L., 1985. Artificial reef research: a review with recommendations for future priorities. Bulletin of Marine Science 11–39. Bord, R.J., O'Connor, R.E., 1997. The gender gap in environmental attitudes: the case of perceived vulnerability to risk: research on the environment. Social Science Quarterly 78, 830–840. Börger, T., Hooper, T.L., Austen, M.C., 2015. Valuation of ecological and amenity impacts of an offshore windfarm as a factor in marine planning. Environmental Science & Policy 54, 126–133. doi:10.1016/j.envsci.2015.05.018 Breukers, S., Wolsink, M., 2007. Wind power implementation in changing institutional landscapes: An international comparison. Energy Policy 35, 2737–2750. doi:10.1016/j.enpol.2006.12.004 Brown, T.C., 1984. The concept of value in resource allocation. Land Economics 60, 231–246. Buhrmester, M., Kwang, T., Gosling, S.D., 2011. Amazon's Mechanical Turk: A new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science 6, 3–5. doi:10.1177/1745691610393980 Burgess, J., Chilvers, J., 2006. Upping the ante: A conceptual framework for designing and evaluating participatory technology assessments. Science and Public Policy 33, 713–728. doi:10.3152/147154306781778551 Burgess, J., Stirling, A., Clark, J., Davies, G., 2007. Deliberative mapping: a novel analytic- deliberative methodology to support contested science-policy decisions. Public Understanding of Science 16, 299–322. doi:10.1177/0963662507077510

163 Campbell, T.H., Kay, A.C., 2014. Solution aversion: On the relation between ideology and motivated disbelief. Journal of personality and social psychology 107, 809–824. doi:10.1037/a0037963 Carlsson, F., Frykblom, P., Johan Lagerkvist, C., 2005. Using cheap talk as a test of validity in choice experiments. Economics Letters 89, 147–152. doi:10.1016/j.econlet.2005.03.010 Cash, D.W., Clark, W.C., Alcock, F., Dickson, N.M., Eckley, N., Guston, D.H., Jäger, J., Mitchell, R.B., 2003. Knowledge systems for sustainable development. Proceedings of the National Academy of Sciences 100, 8086–8091. doi:10.1073/pnas.1231332100 Chan, K., Goldstein, J., Satterfield, T., Hannahs, N., Kikiloi, K., Naidoo, R., Vadeboncoeur, N., Woodside, U., 2011. Cultural services and non-use values, in: Kareiva, P., Tallis, H., Ricketts, T.H., Daily, G.C., Polasky, S. (Eds.), Natural Capital: Theory and Practice of Mapping Ecosystem Services. Oxford University Press, Oxford, pp. 206–228. doi:10.1093/acprof:oso/9780199588992.001.0001/acprof-9780199588992 Chan, K.M.A., Balvanera, P., Benessaiah, K., Chapman, M., Diaz, S., Gómez-Baggethun, E., Gould, R., Hannahs, N., Jax, K., Klain, S., Luck, G.W., Martín-López, B., Muraca, B., norton, B., Ott, K., Pascual, U., Satterfield, T., Tadaki, M., Taggart, J., Turner, N., 2016. Opinion: Why protect nature? Rethinking values and the environment. Proceedings of the National Academy of Sciences 113, 1462–1465. doi:10.1073/pnas.1525002113 Chan, K.M.A., Guerry, A.D., Balvanera, P., Klain, S., 2012a. Where are cultural and social in ecosystem services? A framework for constructive engagement. BioScience 62, 744–756. doi:10.1525/bio.2012.62.8.7 Chan, K.M.A., Pringle, R.M., Ranganathan, J., Boggs, C.L., Chan, Y.L., Ehrlich, P.R., Haff, P.K., Heller, N.E., Khafaji, Al, K., Macmynowski, D.P., 2007. When agendas collide: Human welfare and biological conservation. Conserv. Biol. 21, 59–68. doi:10.1111/j.1523-1739.2006.00570.x Chan, K.M.A., Satterfield, T., Goldstein, J., 2012b. Rethinking ecosystem services to better address and navigate cultural values. Ecological Economics 74, 8–18. doi:10.1016/j.ecolecon.2011.11.011 Cialdini, R.B., Goldstein, N.J., 2004. Social influence: Compliance and conformity. Annu. Rev. Psychol. 55, 591–621. doi:10.1146/annurev.psych.55.090902.142015 Costanza, R., d'Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K., Naeem, S., O'Neill, R.V., Paruelo, J., 1998. The value of the world's ecosystem services and natural capital. Ecological Economics 25, 3–15. Costanza, R., Kubiszewski, I., 2012. The authorship structure of “ecosystem services” as a transdisciplinary field of scholarship. Ecosystem Services 1, 16–25. doi:10.1016/j.ecoser.2012.06.002 Cowell, R., Bristow, G., Munday, M., 2011. Acceptance, acceptability and environmental justice: the role of community benefits in wind energy development. Journal of Environmental Planning and Management 54, 539–557. doi:10.1080/09640568.2010.521047 Crompton, T., Kasser, T., 2010. Human identity: A missing link in environmental campaigning. Environment: Science and Policy for Sustainable Development 52, 23–33. doi:10.1080/00139157.2010.493114 Cronbach, L.J., 1951. Coefficient alpha and the internal structure of tests. Psychometrika 16, 297–334. doi:10.1007/BF02310555 Cummings, R.G., Taylor, L.O., 1999. Unbiased value estimates for environmental goods: a cheap talk design for the contingent valuation method. The American Economic Review 89, 649–665. doi:10.2307/117038 Daily, G., Matson, P., 2008. Ecosystem services: From theory to implementation. Proceedings of the

164 National Academy of Sciences 105, 9455. Daily, G.C., 1997. Nature's Services: Societal Dependence on Natural Ecosystems. Island Press, Washington, DC. Daily, G.C., Polasky, S., Goldstein, J., Kareiva, P.M., Mooney, H.A., Pejchar, L., Ricketts, T.H., Salzman, J., Shallenberger, R., 2009. Ecosystem services in decision making: time to deliver. Frontiers in Ecology and the Environment 7, 21–28. doi:10.1890/080025 Daniel, T.C., Muhar, A., Arnberger, A., Aznar, O., Boyd, J.W., Chan, K.M.A., Costanza, R., Elmqvist, T., Flint, C.G., Gobster, P.H., Gret-Regamey, A., Lave, R., Muhar, S., Penker, M., Ribe, R.G., Schauppenlehner, T., Sikor, T., Soloviy, I., Spierenburg, M., Taczanowska, K., Tam, J., Dunk, von der, A., 2012. Contributions of cultural services to the ecosystem services agenda. Proceedings of the National Academy of Sciences 109, 8812–8819. doi:10.1073/pnas.1114773109 Davidson, R., Duffy, C., Gaze, P., Baxter, A., DuFresne, S., Courtney, S., Hamill, P., 2011. Ecologically significant marine sites in Marlborough, New Zealand. Nelson, NZ. Demski, C., Butler, C., Parkhill, K.A., Spence, A., Pidgeon, N.F., 2015. Public values for energy system change. Global Environmental Change 34, 59–69. doi:10.1016/j.gloenvcha.2015.06.014 Devine-Wright, P., 2011. Place attachment and public acceptance of renewable energy: A tidal energy case study. Journal of Environmental Psychology 31, 336–343. doi:10.1016/j.jenvp.2011.07.001 Devine-Wright, P., 2009. Rethinking NIMBYism: The role of place attachment and place identity in explaining place-protective action. J. Community. Appl. Soc. Psychol. 19, 426–441. doi:10.1002/casp.1004 Devine-Wright, P., 2005. Beyond NIMBYism: towards an integrated framework for understanding public perceptions of wind energy. Wind Energy 8, 125–139. doi:10.1002/we.124 Devine-Wright, P., Howes, Y., 2010. Disruption to place attachment and the protection of restorative environments: A wind energy case study. Journal of Environmental Psychology 30, 271–280. doi:10.1016/j.jenvp.2010.01.008 Devine-Wright, P., Walker, G., Barnett, J., 2011. Renewable Energy and the Public: From “Not in My Backyard” to Participation. Earthscan, London. Diaz, S., Demissew, S., Carabias, J., Joly, C., Lonsdale, M., Ash, N., Larigauderie, A., Adhikari, J.R., Arico, S., Báldi, A., Bartuska, A., Baste, I.A., Bilgin, A., Brondizio, E., Chan, K.M., Figueroa, V.E., Duraiappah, A., Fischer, M., Hill, R., Koetz, T., Leadley, P., Lyver, P., Mace, G.M., Martín- López, B., Okumura, M., Pacheco, D., Pascual, U., Pérez, E.S., Reyers, B., Roth, E., Saito, O., Scholes, R.J., Sharma, N., Tallis, H., Thaman, R., Watson, R., Yahara, T., Hamid, Z.A., Akosim, C., Al-Hafedh, Y., Allahverdiyev, R., Amankwah, E., Asah, S.T., Asfaw, Z., Bartus, G., Brooks, L.A., Caillaux, J., Dalle, G., Darnaedi, D., Driver, A., Erpul, G., Escobar-Eyzaguirre, P., Failler, P., Fouda, A.M.M., Fu, B., Gundimeda, H., Hashimoto, S., Homer, F., Lavorel, S., Lichtenstein, G., Mala, W.A., Mandivenyi, W., Matczak, P., Mbizvo, C., Mehrdadi, M., Metzger, J.P., Mikissa, J.B., Moller, H., Mooney, H.A., Mumby, P., Nagendra, H., Neßhöver, C., Oteng-Yeboah, A.A., Pataki, G., Roué, M., Rubis, J., Schultz, M., Smith, P., Sumaila, R., Takeuchi, K., Thomas, S., Verma, M., Yeo-Chang, Y., Zlatanova, D., 2015. The IPBES conceptual framework — connecting nature and people. Current Opinion in Environmental Sustainability 14, 1–16. doi:10.1016/j.cosust.2014.11.002 Dietz, T., Fitzgerald, A., Shwom, R., 2005. Environmental values. Annu. Rev. Environ. Resourc. 30, 335–372. doi:10.1146/annurev.energy.30.050504.144444 Dietz, T., Stern, P.C., 2008. Public Participation in Environmental Assessment and Decision Making. Panel on Public Participation in Environmental Assessment and Decision Making, National

165 Research Council. Dincer, I., 2000. Renewable energy and sustainable development: A crucial review. Renewable and Sustainable Energy Reviews 4, 157–175. doi:10.1016/S1364-0321(99)00011-8 DOE EIA, 2016. International Energy Outlook 2016. Department of Energy, Washington, DC. DOE EIA, 2015. Wind Vision: A New Era for Wind Power in the United States. Office of Scientific and Technical Information, Oakridge, TN. Douglas, M., Wildavsky, A., 1983. Risk and Culture. University of California Press, Berkeley. Dunlap, R., Liere, K.V., Mertig, A., 2000. Measuring endorsement of the new ecological paradigm: A revised NEP scale. Journal of social issues 56, 425–442. Dunlap, R.E., 2008. The New Environmental Paradigm scale: From marginality to worldwide use. The Journal of Environmental Education 40, 3–18. doi:10.3200/JOEE.40.1.3-18 Dunlap, R.E., McCright, A.M., 2008. A widening gap: Republican and Democratic views on climate change. Environment: Science and Policy for Sustainable Development 50, 26–35. doi:10.3200/ENVT.50.5.26-35 Dunlap, R.E., Van Liere, K.D., 1978. The “New Environmental Paradigm.” The Journal of Environmental Education. doi:10.3200/JOEE.40.1.19-28 Economist, 2015. Wondering about wind. The Economist 1–5. Ehrlich, P.R., Ehrlich, A.H., 1982. Extinction: the causes and consequences of the disappearance of species. Gollancz, London. Ehrlich, P.R., Mooney, H.A., 1983. Extinction, substitution, and ecosystem services. BioScience 33, 248–254. EIA, 2015. Annual Energy Outlook 2015 with projections to 2040. U.S. Energy Information Administration, Washington, D.C. Ek, K., 2002. Valuing the environmental impacts of wind power: A choice experiment approach. Lulea University of Technology. Ek, K., Persson, L., 2014. Wind farms — where and how to place them? A choice experiment approach to measure consumer preferences for characteristics of wind farm establishments in Sweden. Ecological Economics 105, 193–203. doi:10.1016/j.ecolecon.2014.06.001 Eken, G., Bennun, L., Brooks, T.M., Darwall, W., Fishpool, L.D.C., Foster, M., Knox, D., Langhammer, P., Matiku, P., Radford, E., Salaman, P., Sechrest, W., Smith, M.L., Spector, S., Tordoff, A., 2004. Key Biodiversity Areas as Site Conservation Targets. BioScience 54, 1110– 1118. doi:10.1641/0006-3568(2004)054[1110:KBAASC]2.0.CO;2 Entrekin, S., Evans-White, M., Johnson, B., Hagenbuch, E., 2011. Rapid expansion of natural gas development poses a threat to surface waters. Frontiers in Ecology and the Environment 9, 503– 511. doi:10.1890/110053 Epstein, S., 1994. Integration of the cognitive and the psychodynamic unconscious. American Psychologist 49, 709. doi:10.1037/0003-066X.49.8.709 Failing, L., Gregory, R., Harstone, M., 2007. Integrating science and local knowledge in environmental risk management: A decision-focused approach. Ecological Economics 64, 47–60. doi:10.1016/j.ecolecon.2007.03.010 Field, A., Miles, J., Field, Z., 2012. Discovering Statistics Using R. Sage, Los Angeles. Field, J., 2014. UK fishermen offer maine counterparts offshore wind advice, Maine Public Broadcasting Network. Finucane, M.L., Alhakami, A., Slovic, P., Johnson, S.M., 2000a. The affect heuristic in judgments of risks and benefits. J. Behav. Dec. Making 13, 1–17.

166 Finucane, M.L., Slovic, P., Mertz, C., Flynn, J., Satterfield, T.A., 2000b. Gender, race, and perceived risk: The “white male” effect. Health, Risk & Society 2, 159–172. doi:10.1080/713670162 Firestone, J., Archer, C.L., Gardner, M.P., Madsen, J.A., Prasad, A.K., Veron, D.E., 2015. Opinion: The time has come for offshore wind power in the United States. Proceedings of the National Academy of Sciences 201515376–4. doi:10.1073/pnas.1515376112 Firestone, J., Kempton, W., 2007. Public opinion about large offshore wind power: Underlying factors. Energy Policy 35, 1584–1598. doi:10.1016/j.enpol.2006.04.010 Firestone, J., Kempton, W., Krueger, A., 2009. Public acceptance of offshore wind power projects in the USA. Wind Energy 12, 183–202. doi:10.1002/we.316 Firestone, J., Kempton, W., Lilley, M.B., Samoteskul, K., 2012. Public acceptance of offshore wind power across regions and through time. Journal of Environmental Planning and Management 55, 1369–1386. doi:10.1080/09640568.2012.682782 Fortuin, S., Nichol, S., Franz, P., Jamieson, D., Smith, M., Stevens, C., 2009. New Zealand’s EnergyScape. NIWA, Auckland. Foxon, T.J., Hammond, G.P., Pearson, P.J.G., 2010. Developing transition pathways for a low carbon electricity system in the UK. Technological Forecasting & Social Change 77, 1203–1213. doi:10.1016/j.techfore.2010.04.002 Förster, J., Barkmann, J., Fricke, R., Hotes, S., 2015. Assessing ecosystem services for informing land- use decisions: a problem-oriented approach. Ecology and Society. Gee, K., Burkhard, B., 2010. Cultural ecosystem services in the context of offshore wind farming: A case study from the west coast of Schleswig-Holstein. Ecological Complexity 7, 349–358. doi:10.1016/j.ecocom.2010.02.008 Gifford, R., Comeau, L.A., 2011. Message framing influences perceived climate change competence, engagement, and behavioral intentions. Global Environmental Change 21, 1301–1307. doi:10.1016/j.gloenvcha.2011.06.004 Gill, A.B., 2005. Offshore renewable energy: Ecological implications of generating electricity in the coastal zone. Journal of Applied Ecology 42, 605–615. doi:10.1111/j.1365-2664.2005.01060.x Goodale, M.W., Milman, A., 2014. Cumulative adverse effects of offshore wind energy development on wildlife. Journal of Environmental Planning and Management 1–21. doi:10.1080/09640568.2014.973483 Goodman, J.K., Cryder, C.E., Cheema, A., 2012. Data collection in a flat world: the strengths and weaknesses of Mechanical Turk samples. J. Behav. Dec. Making 26, 213–224. doi:10.1002/bdm.1753 Gould, R.K., Klain, S.C., Ardoin, N.M., Satterfield, T., Woodside, U., Hannahs, N., Daily, G.C., Chan, K.M., 2014. A Protocol for Eliciting Nonmaterial Values Through a Cultural Ecosystem Services Frame. Conservation Biology 29, 1–12. doi:10.1111/cobi.12407 Gómez-Baggethun, E., de Groot, R., Lomas, P.L., Montes, C., 2010. The history of ecosystem services in economic theory and practice: From early notions to markets and payment schemes. Ecological Economics 69, 1209–1218. doi:10.1016/j.ecolecon.2009.11.007 Green, R., Vasilakos, N., 2011. The economics of offshore wind. Energy Policy 39, 496–502. doi:10.1016/j.enpol.2010.10.011 Gregory, R., Failing, L., Harstone, M., Long, G., McDaniels, T., Ohlson, D., 2012. Structured Decision Making. John Wiley & Sons. Guadagnoli, E., Velicer, W.F., 1988. Relation of sample size to the stability of component patterns. Psychological Bulletin 103, 265–275.

167 Guerry, A.D., Polasky, S., Lubchenco, J., Chaplin-Kramer, R., Daily, G.C., Griffin, R., Ruckelshaus, M., Bateman, I.J., Duraiappah, A., Elmqvist, T., Feldman, M.W., Folke, C., Hoekstra, J., Kareiva, P.M., Keeler, B.L., Li, S., McKenzie, E., Ouyang, Z., Reyers, B., Ricketts, T.H., Rockström, J., Tallis, H., Vira, B., 2015. Natural capital and ecosystem services informing decisions: From promise to practice. Proceedings of the National Academy of Sciences 112, 7348–7355. doi:10.1073/pnas.1503751112 Habermas, J., 2004. The Theory of Communicative Action. John Wiley & Sons. Haggett, C., 2011. “Planning and pursuasion”: Public engagement in renewable energy decision- making, in: Devine-Wright, P. (Ed.), Renewable Energy and the Public: From “Not in My Backyard” to Participation. Earthscan, London, pp. 15–28. Haidt, J., 2007. The new synthesis in moral psychology. Science 316, 998–1002. doi:10.1126/science.1137651 Haidt, J., 2001. The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychological Review 108, 814. doi:10.1037//0033-295X Halpern, B.S., Walbridge, S., Selkoe, K.A., Kappel, C.V., Wainstein, M., D'agrosa, C., Bruno, J.F., Casey, K.S., Ebert, C., Fox, H.E., Fujita, R., Heinemann, D., Lenihan, H.S., Madin, E.M.P., Perry, M.T., Selig, E.R., Spalding, M., Steneck, R., Watson, R., 2008. A Global Map of Human Impact on Marine Ecosystems. Science 319, 948–952. doi:10.1126/science.1149345 Helgeson, J., van der Linden, S., Chabay, I., 2012. The role of knowledge, learning and mental models in public perceptions of climate change related risks, in: Learning for Sustainability in Times of Accelerating Change. Wageningen Academic Publishers, pp. 1–18. Hoffert, M.I., 2002. Advanced technology paths to global climate stability: Energy for a greenhouse planet. Science 298, 981–987. doi:10.1126/science.1072357 Honey-Rosés, J., Pendleton, L.H., 2013. A demand driven research agenda for ecosystem services. Ecosystem Services Complete, 160–162. doi:10.1016/j.ecoser.2013.04.007 Horton, J.J., Chilton, L.B., 2010. The labor economics of paid crowdsourcing, the 11th ACM conference. ACM, New York, New York, USA. doi:10.1145/1807342.1807376 Huff, C., Tingley, D., 2015. “Who are these people?” Evaluating the demographic characteristics and political preferences of MTurk survey respondents. Research & Politics 2, 2053168015604648. doi:10.1177/2053168015604648 IAN, 2015. IAN Image and Video Library. Integration and Application Network. Inger, R., Attrill, M.J., Bearhop, S., Broderick, A.C., James Grecian, W., Hodgson, D.J., Mills, C., Sheehan, E., Votier, S.C., Witt, M.J., Godley, B.J., 2009. Marine renewable energy: Potential benefits to biodiversity? An urgent call for research. Journal of Applied Ecology. doi:10.1111/j.1365-2664.2009.01697.x IPCC, 2014. Climate Change 2014 Synthesis Report for Policy Makers. Intergovernmental Panel on Climate Change, Geneva, Switzerland. IPCC, 2011. IPCC Special Report on Renewable Energy Sources and Climate Change Mitigation. Cambridge University Press, Cambridge, UK and New York, NY, USA. Irvin, R.A., Stansbury, J., 2004. Citizen participation in decision making: Is it worth the effort? Public Administration Review 64, 55–65. doi:10.1111/j.1540-6210.2004.00346.x Irwin, A., Wynne, B., 2004. Misunderstanding Science? Cambridge University Press. Island Institute, 2015. Ocean Renewable Energy [WWW Document]. httpwww.islandinstitute.orgocean-renewable-energy. URL http://www.islandinstitute.org/ocean- renewable-energy (accessed 12.4.15).

168 Island Institute, 2012a. Offshore Wind Fact Sheet Series for Northeast Coastal Communities [WWW Document]. httpwww.islandedgrid.orgoffshore-wind-fact-sheets. URL http://www.islandedgrid.org/offshore-wind-fact-sheets/ (accessed 8.24.15a). Island Institute, 2012b. Offshore Wind Energy Session at the Maine Fisherman's Forum. Rockland, Maine. Island Institute, 2009. Mapping Working Waters [WWW Document]. httpwww.islandinstitute.orgprogrammarine-programsocean-planningmww. URL http://www.islandinstitute.org/program/marine-programs/ocean-planning#mww (accessed 8.25.15). IUCN, 2010. Greening Blue Energy: Identifying and managing the biodiversity risks and opportunities of offshore renewable energy, IUCN. IUCN. Jacobson, M.Z., Delucchi, M.A., 2011. Providing all global energy with wind, water, and solar power, Part I Technologies, energy resources, quantities and areas of infrastructure, and materials. Energy Policy 39, 1154–1169. doi:10.1016/j.enpol.2010.11.040 Jacobson, M.Z., Delucchi, M.A., Bazouin, G., Bauer, Z.A.F., Heavey, C.C., Fisher, E., Morris, S.B., Piekutowski, D.J.Y., Vencill, T.A., Yeskoo, T.W., 2015a. 100% clean and renewable wind, water, and sunlight (WWS) all-sector energy roadmaps for the 50 United States. Energy & Environmental Science 8, 2093–2117. doi:10.1039/C5EE01283J Jacobson, M.Z., Delucchi, M.A., Cameron, M.A., Frew, B.A., 2015b. Low-cost solution to the grid reliability problem with 100% penetration of intermittent wind, water, and solar for all purposes. Proceedings of the National Academy of Sciences 112, 15060–15065. doi:10.1073/pnas.1510028112 Jax, K., Barton, D.N., Chan, K.M.A., de Groot, R., Doyle, U., Eser, U., Görg, C., Gómez-Baggethun, E., Griewald, Y., Haber, W., Haines-Young, R., Heink, U., Jahn, T., Joosten, H., Kerschbaumer, L., Korn, H., Luck, G.W., Matzdorf, B., Muraca, B., Neßhöver, C., norton, B., Ott, K., Potschin, M., Rauschmayer, F., Haaren, von, C., Wichmann, S., 2013. Ecosystem services and ethics. Ecological Economics 93, 260–268. doi:10.1016/j.ecolecon.2013.06.008 Johansson, T.B., Nakicenovic, N., Patwardhan, A., Gomez-Echeverri, L., 2016. Global Energy Assessment. Cambridge University Press. doi:9780 52118 2935 Kahan, D., 2010. Fixing the communications failure. Nature 463, 296–297. doi:10.1038/463296a Kahan, D.M., 2015. Climate-Science Communication and the Measurement Problem. Political Psychology 36, 1–43. doi:10.1111/pops.12244 Kahan, D.M., 2012. Cultural Cognition as a Conception of the Cultural Theory of Risk, in: Roeser, S. (Ed.), Handbook of Risk Theory. Springer Publishing, pp. 1–47. Kahan, D.M., Peters, E., Wittlin, M., Slovic, P., Ouellette, L.L., Braman, D., Mandel, G., 2012. The polarizing impact of science literacy and numeracy on perceived climate change risks. Nature Climate Change 2, 732–735. doi:10.1038/nclimate1547 Kahneman, D., 2011. Thinking, Fast and Slow. Macmillan, New York. Kareiva, P., Marvier, M., Lalasz, R., 2012. Conservation in the Anthropocene. Breakthrough Journal. Kareiva, P., Tallis, H., Ricketts, T.H., Daily, G.C., Polasky, S., 2011. Natural capital: Theory and practice of mapping ecosystem services. Oxford University Press, Oxford. Karp, D.S., Mendenhall, C.D., Sandí, R.F., Chaumont, N., Ehrlich, P.R., Hadly, E.A., Daily, G.C., 2013. Forest bolsters bird abundance, pest control and coffee yield. Ecology Letters 16, 1339– 1347. doi:10.1111/ele.12173 Kasperson, R.E., Renn, O., Slovic, P., Brown, H.S., Emel, J., Goble, R., Kasperson, J.X., Ratick, S.,

169 1988. The social amplification of risk: A conceptual framework. Risk Analysis 8, 177–187. Keeney, R.L., 2004. Framing public policy decisions. IJTPM 4, 95–22. doi:10.1504/IJTPM.2004.004815 Kempton, W., Firestone, J., Lilley, J., Rouleau, T., Whitaker, P., 2005. The offshore wind power debate: Views from Cape Cod. Coastal Management 33, 119–149. doi:10.1080/08920750590917530 Kimmell, K., Stalenhoef, D.S., 2011. Cape Wind offshore wind energy project: : A case study of the difficult transition to renewable energy. Golden Gate University of Environmental Law Journal. Klain, S., MacDonald, S., Battista, N., 2015. Engaging Communities in Offshore Wind. Island Institute. Klain, S.C., Chan, K.M.A., 2012. Navigating coastal values: Participatory mapping of ecosystem services for spatial planning. Ecological Economics 82, 104–113. doi:10.1016/j.ecolecon.2012.07.008 Klain, S.C., Satterfield, T.A., Chan, K.M.A., 2014. What matters and why? Ecosystem services and their bundled qualities. Ecological Economics 107, 310–320. doi:10.1016/j.ecolecon.2014.09.003 Klein, N., 2014. This Changes Everything. Simon and Schuster, New York. Kremen, C., 2005. Managing ecosystem services: what do we need to know about their ecology? Ecology Letters 8, 468–479. doi:10.1111/j.1461-0248.2005.00751.x Krohn, S., Damborg, S., 1999. On public attitudes towards wind power. Renewable Energy 16, 954– 960. doi:10.1016/S0960-1481(98)00339-5 Krosnick, J.A., MacInnis, B., 2013. Does the American public support legislation to reduce greenhouse gas emissions? Daedalus 142, 26–39. doi:10.1162/DAED_a_00183 Krueger, A., 2007. Valuing public preferences for offshore wind power: A choice experiment approach. University of Delaware. Krueger, A.D., Parsons, G.R., Firestone, J., 2011. Valuing the Visual Disamenity of Offshore Wind Power Projects at Varying Distances from the Shore: An Application on the Delaware Shoreline 87, 268–283. Kuvlesky, W.P., Jr, Brennan, L.A., Morrison, M.L., Boydston, K.K., Ballard, B.M., Bryant, F.C., 2007. Wind energy development and wildlife conservation: Challenges and opportunities. Journal of Wildlife Management 71, 2487–2498. doi:10.2193/2007-248 Ladenburg, J., Dubgaard, A., 2007. Willingness to pay for reduced visual disamenities from offshore wind farms in Denmark. Energy Policy 35, 4059–4071. doi:10.1016/j.enpol.2007.01.023 Leiserowitz, A., 2006. Climate change risk perception and policy preferences: The role of affect, imagery, and values. Climatic Change 77, 45–72. doi:10.1007/s10584-006-9059-9 Levine, J., Chan, K.M.A., Satterfield, T., 2015. From rational actor to efficient complexity manager: Exorcising the ghost of Homo economicus with a unified synthesis of cognition research. Ecological Economics 114, 22–32. doi:10.1016/j.ecolecon.2015.03.010 List, J.A., Gallet, C.A., 2001. What experimental protocol influence disparities between actual and hypothetical stated values? Environmental and Resource Economics 20, 241–254. doi:10.1023/A:1012791822804 Loewenstein, G.F., Weber, E.U., Hsee, C.K., Welch, N., 2001. Risk as feelings. Psychological Bulletin 127, 267–286. doi:10.1037//0033-2909.127.2.267 Loomis, J., 2011. What's to know about hypothetical bias in stated preference valuation studies? Journal of Economic Surveys 25, 363–370. doi:10.1111/j.1467-6419.2010.00675.x Louviere, J.J., Hensher, D.A., Swait, J.D., 2000. Stated Choice Methods. Cambridge University Press,

170 Cambridge. Lyle, J.T., 1996. Regenerative Design for Sustainable Development. John Wiley & Sons, New York. MA, 2003. Millennium Ecosystem Assessment, Ecosystems and Human Well-being: A Framework for Assessment. Island Press, Washington, DC. Maine State Legislature, 2010. An Act To Provide Predictable Benefits to Maine Communities That Host Wind Energy Developments. Second regular session. Martín-López, B., Gómez-Baggethun, E., Lomas, P.L., Montes, C., 2009. Effects of spatial and temporal scales on cultural services valuation. Journal of Environmental Management 90, 1050– 1059. doi:10.1016/j.jenvman.2008.03.013 Martín-López, B., Iniesta-Arandia, I., García-Llorente, M., Palomo, I., Casado-Arzuaga, I., Amo, D.G.D., Gómez-Baggethun, E., Oteros-Rozas, E., Palacios-Agundez, I., Willaarts, B., González, J.A., Santos-Martín, F., Onaindia, M., López-Santiago, C., Montes, C., 2012. Uncovering Ecosystem Service Bundles through Social Preferences. PLoS ONE 7, e38970. doi:10.1371/journal.pone.0038970.t005 Martín-López, B., Montes, C., Benayas, J., 2007. The non-economic motives behind the willingness to pay for biodiversity conservation. Biological Conservation 139, 67–82. doi:10.1016/j.biocon.2007.06.005 Martínez-Harms, M.J., Bryan, B.A., Balvanera, P., Law, E.A., Rhodes, J.R., Possingham, H.P., Wilson, K.A., 2015. Making decisions for managing ecosystem services. Biological Conservation 184, 229–238. doi:10.1016/j.biocon.2015.01.024 Marvier, M., 2013. New conservation: Friend or foe to the traditional paradigm? SNAP.is Magazine 1–12. Marvier, M., Wong, H., 2012. Resurrecting the conservation movement. J Environ Stud Sci 2, 291– 295. doi:10.1007/s13412-012-0096-6 MBIE, 2015. Energy in New Zealand 2015 (No. MB13204), Energy and building trends. Wellington. McCright, A.M., Dunlap, R.E., 2011. The politicization of climate change and polarization in the american public's views of global warming, 2001–2010. The Sociological Quarterly 52, 155–194. McDonough, W., Braungart, M., 2002. Cradle to Cradle. North Point Press, New York. McGlinchey, D., 2013. The cable is the key: Block Islanders debate industrial wind. Island Journal 1– 2. MCP, 2009. Final Report of the Ocean Energy Task Force to Governor John E. Baldacci (No. 020- 07B-008205). Maine Coastal Program, Maine State Planning office, Augusta, Maine. Meadows, D.H., Meadows, D.L., Randers, J., Behrens, W.W., 1972. The Limits to Growth. Universe Books, New York. Milcu, A.I., Hanspach, J., Abson, D., Fischer, J., 2013. Cultural ecosystem services: A literature review and prospects for future research. Ecology and Society 18, 1–44. doi:10.5751/ES-05790- 180344 Milcu, A.I., Sherren, K., Hanspach, J., Abson, D., Fischer, J., 2014. Navigating conflicting landscape aspirations: Application of a photo-based Q-method in Transylvania (Central Romania). Land Use Policy 41, 408–422. doi:10.1016/j.landusepol.2014.06.019 Ministry for the Environment, 2015. New Zealand’s emissions reduction targets [WWW Document]. newzealand.govt.nz. URL (accessed 9.8.15). Mohsen Tavakol, R.D., 2011. Making sense of Cronbach's alpha. International Journal of Medical Education 2, 53–55. doi:10.5116/ijme.4dfb.8dfd Moss, R.H., Edmonds, J.A., Hibbard, K.A., Manning, M.R., Rose, S.K., van Vuuren, D.P., Carter,

171 T.R., Emori, S., Kainuma, M., Kram, T., Meehl, G.A., Mitchell, J.F.B., Nakicenovic, N., Riahi, K., Smith, S.J., Stouffer, R.J., Thomson, A.M., Weyant, J.P., Wilbanks, T.J., 2010. The next generation of scenarios for climate change research and assessment. Nature 463, 747–756. doi:10.1038/nature08823 MPUC, 2015. Delivery Rates, Electricity Statistics. MPUC, 2010. Request for proposals for long-term contracts for deep-water offshore wind energy pilot projects and tidal energy demonstration projects. Maine Public Utilities Commission, Augusta, Maine. Muraca, B., 2011. The map of moral significance: A new axiological matrix for environmental ethics. Environmental Values 20, 375–396. doi:10.3197/096327111X13077055166063 Murphy, J.J., Allen, P.G., Stevens, T.H., Weatherhead, D., 2005. A meta-analysis of hypothetical bias in stated preference valuation. Environmental and Resource Economics 30, 313–325. doi:10.1007/s10640-004-3332-z MVC, 2009. Island Plan. Martha's Vineyard Commission. Naber, H., Jeffrey, N., Starke, L., Brez, J.A., Lecksell, J., Wiederspahn, A.A., Sears, A., 2009. Valuing coastal and marine ecosystem services. Environment Matters at the World Bank 1–56. Neill, H.R., Cummings, R.G., Ganderton, P.T., Harrison, G.W., McGuckin, T., 1994. Hypothetical surveys and real economic commitments 70, 145. doi:10.2307/3146318 Nelson, E., Mendoza, G., Regetz, J., Polasky, S., Tallis, H., Cameron, D., Chan, K.M., Daily, G.C., Goldstein, J., Kareiva, P.M., Lonsdorf, E., Naidoo, R., Ricketts, T.H., Shaw, M., 2009. Modeling multiple ecosystem services, biodiversity conservation, commodity production, and tradeoffs at landscape scales. Frontiers in Ecology and the Environment 7, 4–11. doi:10.1890/080023 Nevin, J.A., 2010. The power of cooperation. The Behavior Analyst 33, 189–191. Nichols, W.J., 2014. Blue Mind. Abacus, New York. Nordhaus, W.D., 2013. The Climate Casino. Yale University Press. Nordlund, A.M., Garvill, J., 2002. Value Structures behind Proenvironmental Behavior. Environment and Behavior 34, 740–756. doi:10.1177/001391602237244 NRC, 1996. Understanding Risk: Informing Decisions in a Democratic Society. Committee on Risk Characterization, National Research Council, Washington, D.C. NREL, 2015. NREL Dynamic Maps, GIS Data, and Analysis Tools - Wind Data. Nutters, H.M., Pinto da Silva, P., 2012. Fishery stakeholder engagement and marine spatial planning: Lessons from the Rhode Island Ocean SAMP and the Massachusetts Ocean Management Plan. Ocean and Coastal Management 67, 9–18. doi:10.1016/j.ocecoaman.2012.05.020 NZME, 2015. Protesters against deep sea oil drilling turn out in force. The New Zealand Herald 1–2. O'Neil, A., 2015. 150 years of news: Wellington's Makara wind farm a clash of eco values. The Dominion Post 1–2. OPS, 2015. Fact Sheet: White House Summit on Offshore Wind 1–4. Ottinger, R.L., Williams, R., 2002. Renewable energy sources for development. Environmental Law 331–368. Paolacci, G., Chandler, J., Ipeirotis, P.G., 2010. Running experiments on Amazon Mechanical Turk. Judgement and Decision Making 5. Pasqualetti, M.J., 2011. Opposing wind energy landscapes: A search for common cause. Annals of the Association of American Geographers 101, 907–917. doi:10.1080/00045608.2011.568879 Peckar, E., 2015a. Vineyard Power Community Engagement. Peckar, E., 2015b. Influence of Fox Island Wind Farm on Vineyard Power.

172 Pelc, R., Fujita, R.M., 2002. Renewable energy from the ocean. Marine Policy 26, 471–479. Pidgeon, N., Demski, C., Butler, C., Parkhill, K., Spence, A., 2014. Creating a national citizen engagement process for energy policy. Proceedings of the National Academy of Sciences 111, 13606–13613. doi:10.1073/pnas.1317512111 Pidgeon, N., Kasperson, R.E., Slovic, P., 2003. The Social Amplification of Risk. Cambridge University Press. Plieninger, T., Dijks, S., Oteros-Rozas, E., Bieling, C., 2013. Assessing, mapping, and quantifying cultural ecosystem services at community level. Land Use Policy 33, 118–129. doi:10.1016/j.landusepol.2012.12.013 Pomeroy, C., Hall-Arber, M., Conway, F., 2014. Power and perspective: Fisheries and the ocean commons beset by demands of development. Marine Policy. doi:10.1016/j.marpol.2014.11.016 Ram, B., 2011. Assessing integrated risks of offshore wind projects: Moving towards gigawatt-scale deployments. Wind Engineering 35, 247–265. doi:10.1260/0309-524X.35.3.247 Randolph, J., Bauer, M., 1999. Improving environmental decision-making through collaborative methods. Policy Studies Review 16, 169–191. doi:10.1111/j.1541-1338.1999.tb00882.x/pdf Raymond, C.M., Singh, G.G., Benessaiah, K., Bernhardt, J.R., Levine, J., Nelson, H., Turner, N.J., Norton, B.G., Tam, J., Chan, K.M.A., 2013. Ecosystem services and beyond: Using multiple metaphors to understand human–environment Relationships. BioScience 63, 536–546. doi:10.1525/bio.2013.63.7.7 Reed, M.S., 2008. Stakeholder participation for environmental management: A literature review. Biological Conservation 2417–2431. Renn, O., 2008. Risk Governance. Taylor & Francis, London. Renn, O., 1999. A Model for an Analytic−Deliberative Process in Risk Management. Environmental Science & Technology. doi:10.1021/es981283m Renn, O., 1992. Risk communication: Towards a rational discourse with the public. Journal of Hazardous Materials 29, 465–519. doi:10.1016/0304-3894(92)85047-5 Renn, O., Klinke, A., van Asselt, M., 2011. Coping with complexity, uncertainty and ambiguity in risk governance: A synthesis. AMBIO: A Journal of the Human Environment 40, 231–246. doi:10.1007/s13280-010-0134-0 Reubens, J.T., Braeckman, U., Vanaverbeke, J., Van Colen, C., Degraer, S., Vincx, M., 2013a. Aggregation at windmill artificial reefs: CPUE of Atlantic cod (Gadus morhua) and pouting (Trisopterus luscus) at different habitats in the Belgian part of the North Sea. Fisheries Research 139, 28–34. doi:10.1016/j.fishres.2012.10.011 Reubens, J.T., Vandendriessche, S., Zenner, A.N., Degraer, S., Vincx, M., 2013b. Offshore wind farms as productive sites or ecological traps for gadoid fishes? - Impact on growth, condition index and diet composition. Marine Environmental Research 1–9. doi:10.1016/j.marenvres.2013.05.013 RNZ, 2012a. Meridian drops West Coast hydro dam plan, Radio New Zealand. RNZ, 2012b. Meridian pulls plug on Central Otago wind farm, Radio New Zealand. Roberts, T., Upham, P., McLachlan, C., Mander, S., Gough, C., Boucher, P., Ghanem, D.A., 2013. Low-Carbon Energy Controversies. Routledge, London and New York. Robinson, J., Cole, R.J., 2014. Theoretical underpinnings of regenerative sustainability. Building Research & Information 43, 133–143. doi:10.1080/09613218.2014.979082 Rockström, J., Steffen, W., Persson, A., Chapin, F.S.I., Lambin, E., Lenton, T.M., Scheffer, M., Folke, C., Schellnhuber, H.J., Nykvist, B., Foley, J., de Wit, C.A., Hughes, T., van der Leeuw, S., Rodhe,

173 H., Sorlin, S., Snyder, P.K., Costanza, R., Svedin, U., Falkenmark, M., Karlberg, L., Corell, R.W., Fabry, V.J., Hansen, J., Walker, B., Liverman, D., Richardson, K., Crutzen, P., 2009. Planetary boundaries: Exploring the safe operating space for humanity. Ecology and Society 14, 32. Ruckelshaus, M., McKenzie, E., Tallis, H., Guerry, A., Daily, G., Kareiva, P., Polasky, S., Ricketts, T., Bhagabati, N., Wood, S.A., Bernhardt, J., 2013. Notes from the field: Lessons learned from using ecosystem service approaches to inform real-world decisions. Ecological Economics. Rudolph, D., Haggett, C., Aitken, M., 2015. Community Benefits from Offshore Renewables: Good Practice Review, ClimateXChange. Russell, R., Guerry, A.D., Balvanera, P., Gould, R.K., Basurto, X., Chan, K.M.A., Klain, S., Levine, J., Tam, J., 2013. Humans and Nature: How Knowing and Experiencing Nature Affect Well- Being. Annu. Rev. Environ. Resourc. 38, 473–502. doi:10.1146/annurev-environ-012312-110838 Ryan, R.M., Deci, E.L., 2001. On happiness and human potentials: A review of research on hedonic and eudaimonic well-being. Annu. Rev. Psychol. doi:10.1146/annurev.psych.52.1.141 Ryfe, D.M., 2005. Does deliberative democracy work? Annu. Rev. Polit. Sci. 8, 49–71. doi:10.1146/annurev.polisci.8.032904.154633 Ryff, C.D., Singer, B.H., 2008. Know Thyself and Become What You Are: A Eudaimonic Approach to Psychological Well-Being. J Happiness Stud 9, 13–39. doi:10.1007/s10902-006-9019-0 SAB, 2009. Valuing the Protection of Ecological Systems and Services: A report of the Science Advisory Board of the U.S. Environmental Protection Agency. Environmental Protection Agency, Washington, DC. Sandel, M.J., 2012. What Money Can't Buy. Penguin UK. Satterfield, T., Conti, J., Harthorn, B.H., Pidgeon, N., Pitts, A., 2012. Understanding shifting perceptions of nanotechnologies and their implications for policy dialogues about emerging technologies. Science and Public Policy 40, 247–260. doi:10.1093/scipol/scs084 Satterfield, T., Kandlikar, M., Beaudrie, C.E.H., Conti, J., Harthorn, B.H., 2009. Anticipating the perceived risk of nanotechnologies. Nature Nanotechnology 4, 752–758. doi:10.1038/nnano.2009.265 Satterfield, T.A., Mertz, C.K., Slovic, P., 2004. Discrimination, vulnerability, and justice in the face of risk. Risk Anal. 24, 115–129. doi:10.1111/j.0272-4332.2004.00416.x Seppelt, R., Dormann, C.F., Eppink, F.V., Lautenbach, S., Schmidt, S., 2011. A quantitative review of ecosystem service studies: approaches, shortcomings and the road ahead. Journal of Applied Ecology 48, 630–636. doi:10.1111/j.1365-2664.2010.01952.x Shellenberger, M., Nordhaus, T., 2009. Break Through. Houghton Mifflin Harcourt, Boston, New York. Shellenberger, M., Nordhaus, T., 2004. The death of environmentalism. The Breakthrough Institute. Sheppard, S., Cizek, P., 2009. The ethics of Google Earth: Crossing thresholds from spatial data to landscape visualisation. Journal of Environmental Management. Sherren, K., Fischer, J., Clayton, H., Schirmer, J., Dovers, S., 2010. Integration by case, place and process: transdisciplinary research for sustainable grazing in the Lachlan River catchment, Australia. Landscape Ecology 25, 1219–1230. doi:10.1007/s10980-010-9494-x Sloman, S.A., 1996. The empirical case for two systems of reasoning. Psychological Bulletin 119, 3– 22. doi:10.1037/0033-2909.119.1.3 Slovic, P., 2010. Risk as analysis and risk as feelings: Some thoughts about affect, reason, risk and rationality, in: The Feeling of Risk: New Perspectives on Risk Perception. Earthscan, New York, p. 456.

174 Slovic, P., 1999. Trust, emotion, sex, politics, and science: Surveying the risk-assessment battlefield. Risk Analysis 19, 689–701. Slovic, P., 1987. Perception of risk. Science 236, 280–285. Slovic, P., Finucane, M.L., Peters, E., MacGregor, D.G., 2007. The affect heuristic. European Journal of Operational Research 177, 1333–1352. doi:10.1016/j.ejor.2005.04.006 Slovic, P., Peters, E., 2006. Risk perception and affect. Current Directions in Psychol Sci 15, 322–325. doi:10.1111/j.1467-8721.2006.00461.x Slovic, S., Slovic, P., 2010. Numbers and nerves: toward an affective apprehension of environmental risk, in: The Feeling of Risk: New Perspectives on Risk Perception. Routledge, New York, pp. 1– 1. Smith, A., Stehly, T., Musial, W., 2015. 2014-2015 Offshore Wind Technologies Market Report, National Renewable Energy Laboratory. NREL, Oakridge, TN. Snyder, B., Kaiser, M.J., 2009. Ecological and economic cost-benefit analysis of offshore wind energy. Renewable Energy 34, 1567–1578. doi:10.1016/j.renene.2008.11.015 Soulé, M., 2013. The “New Conservation.” Conservation Biology 27, 895–897. doi:10.1111/cobi.12147 Sovacool, B.K., 2014. What are we doing here? Analyzing fifteen years of energy scholarship and proposing a social science research agenda. Energy Policy 1, 1–29. doi:10.1016/S0301- 4215(99)00012-9 Spash, C.L., 2008a. Deliberative monetary valuation and the evidence for a new value theory. Land Economics 84, 469–488. Spash, C.L., 2008b. How much is that ecosystem in the window? The one with the bio-diverse trail. Environmental Values 17, 259–284. doi:10.3197/096327108X303882 Steffen, W., Richardson, K., Rockstrom, J., Cornell, S.E., Fetzer, I., Bennett, E.M., Biggs, R., Carpenter, S.R., de Vries, W., de Wit, C.A., Folke, C., Gerten, D., Heinke, J., Mace, G.M., Persson, L.M., Ramanathan, V., Reyers, B., Sorlin, S., 2015. Planetary boundaries: Guiding human development on a changing planet. Science. doi:10.1126/science.1259855 Stehr, N., 2016. Exceptional circumstances: Does climate change trump democracy? Issues in Science and Technology. Stephenson, J., Lawson, R., 2013. Giving voice to the “silent majority.” Policy Quarterly 1–8. Stern, P.C., Dietz, T., Abel, T.D., Guagnano, G.A., 1999. A value-belief-norm theory of support for social movements: The case of environmentalism. Human Ecology Review 6. Stirling, A., 2008. "Opening up" and “closing down”: Power, participation, and pluralism in the social appraisal of technology. Science, Technology & Human Values 33, 262–294. doi:10.1177/0162243907311265 Tallis, H., Levin, P.S., Ruckelshaus, M., Lester, S.E., Mcleod, K.L., Fluharty, D.L., Halpern, B.S., 2010. The many faces of ecosystem-based management: Making the process work today in real places. Marine Policy 34, 340–348. doi:10.1016/j.marpol.2009.08.003 Tallis, H., Lubchenco, J., 2014. A call for inclusive conservation. Nature 515, 27–28. Tallis, H., Polasky, S., 2011. Assessing multiple ecosystem services: an integrated tool for the real world, in: Kareiva, P., Tallis, H., Daily, G.C., Ricketts, T., Polasky, S. (Eds.), Natural Capital: Theory and Practice of Mapping Ecosystem Services. Oxford University Press, pp. 34–50. Tashakkori, A., Teddlie, C., 2003. Handbook of mixed methods in social and behavioral research. Sage Publications, London, UK. Tercek, M.R., Adams, J.S., 2013. Nature's Fortune. Perseus Books Group, New York.

175 Tobias, T.N., 2009. Living Proof. Union of BC Indian Chiefs and Ecotrust Canada, Vancouver, BC. Toke, D., 2002. Wind power in UK and Denmark: Can rational choice help explain different outcomes? Environmental Politics 11, 83–100. doi:10.1080/714000647 Turkel, T., 2016. UMaine wind power project back in running for major federal grant. Portland Press Herald 1–7. Turkel, T., 2015. Maine offshore wind project still faces money hurdles, despite federal grant. Portland Press Herald 1–6. Tversky, A., Kahneman, D., 1974. Judgment under uncertainty: Heuristics and biases. Science 185, 1124–1131. doi:10.1126/science.185.4157.1124 U.S. Census, 2010. 2010 Census Data. UN, 2015. Transforming Our World: the 2030 Agenda for Sustainable Development (No. A/RES/70/1). United Nations. VPC, 2015. Vineyard Power Cooperative [WWW Document]. httpwww.vineyardpower.com. URL (accessed 8.22.15). VPC, OMW, 2015. Community Benefits Agreement Summary, Vineyard Power Cooperative and Offshore MW. Vineyard Power Cooperative. Vucetich, J.A., Bruskotter, J.T., Nelson, M.P., 2015. Evaluating whether nature's intrinsic value is an axiom of or anathema to conservation. Conservation Biology 29, 321–332. doi:10.1111/cobi.12464 Wagner, H.-J., Baack, C., Eickelkamp, T., Epe, A., Lohmann, J., Troy, S., 2011. Life cycle assessment of the offshore wind farm alpha ventus. Energy 36, 2459–2464. doi:10.1016/j.energy.2011.01.036 Walker, B.J.A., Wiersma, B., Bailey, E., 2014. Community benefits, framing and the social acceptance of offshore wind farms: An experimental study in England. Energy Research & Social Science 3, 46–54. doi:10.1016/j.erss.2014.07.003 Walker, G., 1995. Renewable energy and the public. Land Use Policy 12, 49–59. Warren, C.R., Lumsden, C., O'Dowd, S., Birnie, R.V., 2005. “Green On Green”: Public perceptions of wind power in Scotland and Ireland. Journal of Environmental Planning and Management 48, 853–875. doi:10.1080/09640560500294376 Warren, C.R., McFadyen, M., 2010. Does community ownership affect public attitudes to wind energy? A case study from south-west Scotland. Land Use Policy 27, 204–213. doi:10.1016/j.landusepol.2008.12.010 Watling, L., Norse, E.A., 1998. Disturbance of the seabed by mobile fishing gear: a comparison to forest clearcutting. Conservation Biology 12, 1180–1197. doi:10.1046/j.1523- 1739.1998.0120061180.x Webler, T., Tuler, S., Dow, K., Whitehead, J., Kettle, N., 2014. Design and evaluation of a local analytic-deliberative process for climate adaptation planning. Local Environment. doi:10.1080/13549839.2014.930425 WEF, 2011. Scaling up renewables. Geneva. Weisser, D., 2007. A guide to life-cycle greenhouse gas (GHG) emissions from electric supply technologies. Energy 32, 1543–1559. doi:10.1016/j.energy.2007.01.008 Westerberg, V., Jacobsen, J.B., Lifran, R., 2013. The case for offshore wind farms, artificial reefs and sustainable tourism in the French mediterranean. Tourism Management 34, 172–183. doi:10.1016/j.tourman.2012.04.008 Wiersma, B., 2016. Public acceptability of offshore renewable energy in Guernsey: Using visual methods to investigate local energy deliberations. Exeter.

176 Wiersma, B., Devine-Wright, P., 2014. Public engagement with offshore renewable energy: A critical review. WIREs Climate Change 5, 493–507. doi:10.1002/wcc.282 Wilsdon, J., Willis, R., 2004. See-through Science. Demos, London. Wolk, R.M., 2008. Utilizing Google Earth and Google Sketchup to visualize wind farms, in:. Presented at the IEEE International Symposium on Technology and Soceity, IEEE, pp. 1–8. Wolsink, M., 2010. Near-shore wind power—Protected seascapes, environmentalists’ attitudes, and the technocratic planning perspective. Land Use Policy 27, 195–203. doi:10.1016/j.landusepol.2009.04.004 Wolsink, M., 2007. Planning of renewables schemes: Deliberative and fair decision-making on landscape issues instead of reproachful accusations of non-cooperation. Energy Policy 35, 2692– 2704. doi:10.1016/j.enpol.2006.12.002 Wolsink, M., 2000. Wind power and the NIMBY-myth: Institutional capacity and the limited significance of public support. Renewable Energy 21, 49–64. Wüstenhagen, R., Wolsink, M., Bürer, M.J., 2007. Social acceptance of renewable energy innovation: An introduction to the concept. Energy Policy 35, 2683–2691. doi:10.1016/j.enpol.2006.12.001 Wynne, B., 1992. Misunderstood misunderstanding: Social identities and public uptake of science. Public Understand Science 1, 281–304. doi:10.1088/0963-6625/1/3/004 Wynne, B., 1989. Sheepfarming after Chernobyl: A case study in communicating scientific information. Environment: Science and Policy for Sustainable Development 31, 10–39. Yergin, D., 2011. The Quest. Penguin. Zografos, C., Martínez-Alier, J., 2009. The politics of landscape value: a case study of wind farm conflict in rural Catalonia. Environ. Plann. A 41, 1726–1744. doi:10.1068/a41208

~~177 Appendices~~

~~Appendix A Golden Bay interview consent form~~

~~University of British Columbia Institute for Resources, Environment & Sustainability 2202 Main Mall Vancouver, BC Canada V6T 1Z4 Tel: 604.822.7725 Fax: 604.822.9250~~

~~www.ires.ubc.ca~~

~~Consent Form: Exploring Perspectives on Energy and the Environment in Golden Bay~~

~~To:~~

~~Principal Investigator Co-Investigator Dr. Kai Chan Sarah Klain~~

~~This research will contribute towards Sarah Klain’s PhD dissertation.~~

Purpose You are invited to take part in this research because of your professional expertise and/or community leadership. The purpose of this project is to collect information from a wide range of experts and potential stakeholders to better understand attitudes towards energy security, a hypothetical renewable energy project and its potential environmental impacts.

Sponsor This project was made possible by a research grant from the New Zealand Ministry of Business Innovation and Employment. The University of British Columbia is conducting this study in collaboration with Cawthron Institute.

Study Procedures Participating in this study entails an interview that will last approximately 45 minutes to one hour. You will be asked questions about if and how you identify with Golden Bay and the wider region. The interview will also include questions about energy security and a hypothetical offshore wind farm, which will include an interactive visualization of a wind farm in Golden Bay. With your consent, the interview will be audio recorded. After the interview, the digital audio recording will be transcribed and the original files will be deleted to protect confidentiality.

Potential Risks The topics of this interview, including energy security and a hypothetical change in infrastructure, may be contentious. To minimize and avoid psychological stress, the confidentiality of the information that you share is guaranteed and you are free to stop participating in the interview at any point.

178 To minimise the risk of accidental release of confidential information, we will code all interview data and delete the original audio files. Only aggregated data and information that does not reveal the identity of any participant will be published and presented publically.

Potential Benefits Your participation in this study will help researchers and management agencies understand local and regional concerns regarding energy security and potential developments in this area. Through this research we hope to communicate the diversity of values and opinions associated with a hypothetical change to Golden Bay. As someone whose profession involves direct work with ecosystems, energy and/or your community, we feel it is important to include your perspective in this research. If you are interested in receiving a digital copy of the output of this research, please email Sarah Klain at XXX.

Confidentiality Your identity and participation in this research will be kept strictly confidential. All notes and digital audio recordings will be coded and stored on an external hard drive that will be kept in a locked file cabinet. Participants will not be identified by name in project reports.

~~Remuneration/Compensation To thank you for your participation, you will be entered in a draw to win a $75 gift certificate redeemable at any FreshChoice supermarket.~~

~~Contact for information about the study If you have questions or want to know more information about this study, please call or email Sarah Klain at XXX.~~

Contact for concerns about the rights of research subjects If you have any concerns about your treatment or rights as a research subject, you may contact the Research Subject Information Line in the UBC Office of Research Services at 604-822-8598 or if long distance e-mail to [email protected].

~~Consent Your participation in this study is entirely voluntary and you may refuse to participate or withdraw from the study at any time.~~

~~Your signature below indicates that you have received a copy of this consent form for your own records. Your signature also indicates that you consent to participate in this study.~~

~~______Subject Signature Date~~

~~179~~

~~Appendix B Golden Bay Interview request letter~~

~~Exploring Perspectives on Energy and the Environment in Golden Bay~~

~~Dear~~

Based on your expertise and experience, you are invited to take part in a research project to better understand perspectives on energy security and environmental impacts of energy-related developments in Golden Bay. Provided you are willing to take part in this study, you will be asked question about if and how you identify with Golden Bay and the wider region. The interview will also include questions about energy security and an interactive visualization of a wind farm in Golden Bay.

Participating in this study entails an interview that will last approximately 45 minutes. Only aggregated data and information that does not reveal the identity of any participant will be published and presented publically.

This research project is made possible by a grant from the New Zealand Ministry of Business Innovation and Employment. Researchers at the University of British Columbia in Canada are conducting this study in conjunction with Cawthron Institute, based in Nelson as part of a wider study on Tasman and Golden Bay. The purpose of this project is to collect information from a wide range of experts and potential stakeholders to better understand attitudes towards energy security, a hypothetical renewable energy project and its potential environmental impacts. As someone who is involved with decision that effect ecosystems, energy and/or your community, we feel it is important to include your perspective in this research.

~~To thank you for your participation, you will be entered in a draw to win a $75 gift certificate redeemable at any FreshChoice supermarket.~~

If you are able to make time for an interview between 8 and 25 April, please contact Sarah Klain (tel XXX). Also, if you wish to obtain a digital copy of the output of this research, please email Sarah Klain.

~~Sincerely,~~

180 Sarah Klain Principal Investigator PhD Student Dr. Kai Chan Institute for Resources, Environment and Institute for Resources, Environment and Sustainability Sustainability University of British Columbia University of British Columbia

~~This research will contribute towards Sarah Klain's PhD dissertation~~

~~181~~

~~Appendix C Golden Bay interview protocol~~

~~Interviewee #______Date ______Exploring Perspectives on Electricity and the Environment in Golden Bay~~

~~Interview Protocol~~

Introduction Introduce yourself and the project. Thank the participant in advance. Provide: Project description Study region is Golden and Tasman Bay Overview of interview A reminder that this is an exploration and there are no right or wrong answers Consent form and confidentiality agreement

Start the digital recording device. Participant information What is your name? What year were you born? What town do you live in? Can you tell me about your current occupation? Initial Ranking The topic of this interview may seem far from your area of expertise but bear with me. I’m interviewing you based on your experience in [her/his line of work].

182 When thinking about new sources of electricity, which of the following concerns are most important? Please choose your top four concerns, ranking these from most important (1=most important, 2= 2nd most important, 3=3rd most important, 4=4th most important).

~~[Display cards randomly, i.e., don’t use same order every time. Then ask them to sort their top four concerns from high to low, which you can then record as a number above the “__”]~~

__Using more local resources to generate electricity rather than imported resources __Minimizing capital cost of the technology (the one-time cost of the new infrastructure) __Ensuring that utility bills don’t increase more than 10% to cover new costs __Prioritize low carbon source of electricity __Reduce or mitigate strongly any local environmental impacts __Ensure any visual or aesthetic impacts of energy infrastructures are locally acceptable __Ensure noise associated with electricity generation is locally acceptable __ Ability for the energy system to withstand or recover quickly from natural hazards, e.g., an earthquake or storm events __ Other, specify ______

[If people are really struggling with choosing, reduce the list by removing the 1st (local) and 4th (low carbon) items] Can you say more about why [x,y,z] are your top 3 concerns? New Zealand’s Electricity New Zealand gets most of its electricity from hydroelectric dams but heavily relies on fossil fuels for transportation. [show and explain graphics]

~~Figure 1. New Zealand’s Electricity Sources. Hydroelectric dams provide over half of New Zealand’s electricity.~~

~~Oil 0% Coal Gas 7% 17% Wind 5% Biogas Hydro 1% 56% Geo- thermal 14%~~

183 Figure 2. New Zealand’s Total Primary Energy Supply (TPES). TPES is the sum of domesGc producGon of energy plus imported sources of energy, subtracGng energy exports and energy used for internaGonal transport. Primary energy can be renewable or non-re

~~Gas Hydro 19% 11%~~

~~Geothermal 20%~~

~~Oil 34% Coal Other 7% Renewables 9%~~

~~Figure 3. New Zealand’s Consumer Energy Demand by Sector. The transportaGon sector heavily relies on oil. Electricity 250 Renewables 200 Natural Gas Oil 150 Coal 100 Petajoules 50~~

0 Agriculture, Industrial Commercial Transport ResidenGal Forestry and Fishing Sector Electrifying the transportation sector could reduce carbon emissions. This would entail developing additional sources of low carbon electricity.

~~184 Energy Security Now I’d like to talk about energy in this region, in particular at the top of the South Island.~~

~~Let’s begin with the term energy security. What does energy security mean to you? What other words or terms come to mind when you hear the phrase ‘energy security’.~~

I’d like to read one widely accepted definition of the term energy security so that I can be sure that we’re talking about the same thing. Energy security is based on the extent to which energy sources are available, accessible, affordable and acceptable. This includes safety of energy fuels and services, energy efficiency, diversification of supply and minimization of price volatility.

~~Is that an acceptable definition or would you like to offer another one? Yes ____ No ____~~

~~[if no, Other definition ______]~~

~~On a scale of 1 (not secure) to 5 (highly secure), how energy secure do you think this region is?~~

~~[If appropriate] What would a more energy secure top of the south look like?~~

~~Do you think energy security is playing a significant role in how this region is developing?~~

What, if any role, should energy security play in how this region develops? Renewable Electricity Sources The next questions are about renewable sources of electricity, defined as electricity from natural resources that are continuously replenished. Examples include solar power, wind power, hydropower, geothermal and biomass.

~~To what extent do you agree with the following statements [show scale]:~~

~~1 ------2------3------4------5 strongly disagree disagree neutral agree strongly agree~~

We need more development of renewable electricity nationally. We need more development of renewable electricity regionally. Can you elaborate on your answers? What best explains why you agree/disagree/feel neutral?

~~Are there particular renewable electricity projects that you support? Why? Are there particular renewable electricity projects that you oppose? Why?~~

~~Here’s a map of existing and proposed land-based wind farms.~~

~~185 Land-Based Wind Farms ! Proposed Awhitu Hauauru Ma Raki ! Operating or under construction Taharoa Taumatatotara~~

~~Hawkes Bay Titiokura Central Wind~~

~~Waitahora Turitea Mt Munro Puketoi Castle Hill~~

~~New Zealand Wind Energy Association Long Gully~~

~~Lake Grassmere~~

~~Hurunui Mt Cass~~

~~Mt Stalker~~

~~Kaiwera Downs Slopedown Flat Hill ±~~

~~0 25 50 100 150 200 km~~

~~Proposed and existing wind farms in New Zealand.~~

~~Do you have an opinion about any of the land-based wind farms that have been proposed and/or built in New Zealand?~~

Do you: strongly disagree, disagree, neutral, agree, strongly agree with the following statement? New Zealand should prioritize the development of renewable electricity sources other than hydroelectric dams. 1 ------2------3------4------5 strongly disagree disagree neutral agree strongly agree Earthquake Risk Anyone living on the south island is no stranger to earthquake risks. Geologists estimate a 30% chance that the Alpine Fault will rupture in the next 50 years, producing a large earthquake in the range of the major Christchurch quakes in 2010 and 2011. Conceivably, this could significantly disrupt the electricity supply to the region. [Show map]

186 Power Stations & Fault Lines South Island, New Zealand Power Station Capacity Faults Fossil Fuel (MW) Transmission lines (kV) 0 - 51 11 52 - 140 33 141 - 264 50 265 - 500 66 501 - 850 110 Hydroelectric (MW) 220 0 - 51 350 52 - 140 141 - 264 265 - 500 501 - 850

~~0 15 30 60 90 120 km~~

~~South Island fault zones, powerlines and power stations.~~

Because the electricity supply to the top of the south Island is only a few transmission lines that run near fault lines, an earthquake of this magnitude would likely have major consequences for the energy grid.

~~If local renewable energy development significantly reduced this vulnerability, would you [increase/decrease/not change] your support of localized energy development? Discuss.~~

~~In general, did this map influence how you think about energy security? Please explain.~~

~~Is the map surprising? Please explain.~~

~~Offshore Wind Energy Next I’d like to talk about electricity generated by offshore winds. Have you heard of offshore wind farms?~~

[If even vague yes] Can you describe for me any impressions or ideas of what this is and how you think it works? What words or terms come to mind when you hear offshore wind farm? What are your general concerns about developing offshore wind farms? What do you think are the benefits of developing offshore wind?

~~Place Attachment & Identity Now I want to get you to think about the region where you live. Are you attached to the Golden Bay/Tasman Bay region? If so, to what extent are you attached?~~

~~187 1 ------2------3------4------5 unattached mildly attached moderately attached strongly attached Why?~~

Is being ‘from’ Golden/Tasman Bay important to your sense of ‘who you are’ or ‘where you belong.’ 1 ------2------3------4------5 Strongly No No neutral yes strongly yes Can you tell me why? How long have you lived in this area? This region of Tasman and Golden Bay? Are there places on the landscape or seashore here that you are particularly attached to? Can you name these and/or tell me why they matter to you?

~~Visualization Video Now we’re now going to turn our attention to a video about this area and a hypothetical offshore wind farm. [Show movie]~~

[Audio in movie] New Zealand has exceptional wind resources. Golden Bay is one of the two best sites in the country for an offshore wind farm due to its relatively shallow water depths and consistent wind.

The location of this hypothetical wind farm is based on a study conducted by the National Institute of Water and Atmospheric Research. This study identified where wind conditions are suitable for this technology.

The hypothetical farm was placed south of Farewell Spit because of the strong and consistent wind where the water depth is less than 30m. Wind farms at depths greater than 30m are considerably more expensive.

~~Sailing tends to be permitted near offshore wind farms.~~

~~This type of turbine is used in the Horns Rev wind farm off the coast of Denmark. The turbine height from the tip of the blade to sea level is 110 meters. Each turbine blade is 40m in length.~~

This view is from the beach near the Farewell Spit Café and Visitor Centre. Golden Bay is important bird habitat. Research in Denmark showed that geese and ducks altered their flight behavior after a wind farm was built to avoid colliding with the turbines. Similar studies have not yet been done on seabirds in New Zealand.

~~On a clear day, boaters launching near Pohara may see the turbines on the distant horizon.~~

~~People in some northern coastal areas of Able Tasman National Park may be able to see the farm in the distance during clear weather conditions.~~

~~Sound associated with the construction of offshore wind farms can disturb marine mammals.~~

The underwater foundations of the turbines may benefit marine mammals and other sea life because of the habitat they create, which is called the artificial reef effect. An artificial reef is a human-made underwater structure. Algae and invertebrates, such as barnacles and oysters, attach to the hard surfaces of an artificial reef. This marine life can provide habitat and food for fish and other species.

~~188 Bottom trawling would not be allowed within an offshore wind farm or near the underwater electricity cables linking the wind farm to land.~~

~~Recreational fishing may benefit from the artificial reef effect associated with the bases of the turbines.~~

To give you an idea of how much energy can be generated from a wind farm, a 2 MW offshore turbine in a place with consistent, strong wind can power approximately 800 households. This hypothetical wind farm of 25 turbines could power ~20,000 households, which is more than the number of households in the Tasman District.

~~An offshore wind farm in Golden Bay would be a source of electricity with minimal carbon emissions.~~

[when the narration stops you can ask the following] Is there any part of the video you would like to see again? How does this visualization make you feel? [probe this] Why? What does the visualization make you think about? Do you think this technology would cause issues?

Ecosystem Services There are commercial and recreational fisheries, recreational boating and aquaculture in this Bay. This area is also habitat to marine mammals, birds and other species. Earlier you mentioned [XXX] as [natural features or special places] that contribute to your attachment to this region.

~~If you think about the ways in which nature and this place is important to you, what do you think could be lost if this project was developed?~~

~~What, if any, impact would it have on your livelihood?~~

~~What do you think could be gained if it went through?~~

~~Opinions on offshore wind On a scale of 1(strongly disagree) to 5(strongly agree), what do you think of the following statement?~~

~~Offshore wind farms are a promising technology for Golden Bay. 1 ------2------3------4------5 strongly disagree disagree neutral agree strongly agree~~

~~Why?~~

[Give interviewee 20 tokens] Using these 20 tokens, can you weight your concerns related to a potential offshore wind development in Golden Bay? Please assign a higher number of tokens to the issues that you are most worried about.

~~189~~

~~Why did you weight [X] the most? Why did you weight [Y] the least?~~

[X] and [Y] are your top two concerns. Assuming an offshore wind farm was to proceed, in order to ensure these impacts are mostly eliminated, would you be willing to increase your annual income tax burden by: [circle] 1. $0 2. $50 3. $100 4. $150 5. $200 Can you weight the benefits that you associate with a potential offshore wind farm using 20 tokens?

~~190~~

~~Why did you weight [X] the most? Why did you weight [Y] the least?~~

[X] and [Y] are your top two benefits. Assuming an offshore wind farm was to proceed, in order to ensure these benefits are achieved, would you be willing to increase your annual income tax burden by: [circle] 1. $0 2. $50 3. $100 4. $150 5. $200 Final Ranking I’d like to return to one of my initial questions. When thinking about new sources of electricity in general, after this interview, would you change how you rank the following? Again, please choose your top four concerns, ranking these from most important to lesser importance.

~~191 __ Ability for the energy system to withstand or recover quickly from natural hazards, e.g., an earthquake or storm events __ Other, specify ______~~

~~[if different from initial ranking] Can you tell me why you have a different ranking?~~

~~Let me know if you have any additional comments or questions about this interview. Thank-you for your time. [stop recording device]~~

~~192~~

Appendix D Full table of risk components Scoring of risk characteristics from psychometric risk paradigm associated with perceived risks to ecosystem services from an offshore wind farm. Our scores in blue are based on reviewing the biological and social science literature on offshore wind farms as well as interviews conducted in our study site. Our risk characteristics (components of Factor 1 and 2) were inspired by the psychometric risk paradigm. Italics denote components removed from correlation test because the scores to not vary across the potential ES consequences (-) --> diminishes risk perception; (+) --> increases risk perception; WF is wind farm Factor 1 Dread Do people have Do those benefiting bear their Can the person who suffers Can precautions be easily Does this pose a risk Do the risks a choice in Risk factor Does potential consequence Is a particular consequence share of risks? Are risks and Is the potential scale of negative consequences control taken to reduce the negative to people in the increase over exposing question evoke a feeling of dread? fatal? benefits equitably distributed the consequence global? the severity of the consequences? impact? future? time? themselves to across society? this risk? Diminishes risk Consequences not fatal Not globally Low risk to future Risk decreasing Voluntary Controllable (-) Not dread (-) Easily reduced (-) Equitable (-) perception (-) catastrophic (-) generations (-) over time (-) exposure (-) Car: driver can drive cautiously to Risk of cancer Medical x-ray: wear a lead car: drivers benefit from cars Skiing, Example reduce severity of potential Bicycle, car Medical x-ray Fires, floods Medical x-rays after quitting apron, bicycle: wear a helmet while facing risks of driving skydiving accident smoking Globally catastrophic High risk to future Risk increasing Involuntary Increases risk Uncontrollable (+) Dread (+) Consequences Fatal (+) Not easily reduced (+) Not Equitable (+) perception (+) generations (+) over time (+) exposure + Sea level rise due to climate Exposure to Airplane: passengers relinquish Terrorism, shark attack, nuclear change: poor people, who emit pollutants often Example control to pilot, passengers do Nuclear meltdown Ocean acidification Nuclear meltdown Climate change Nuclear fall out meltdown less carbon, will suffer most increase health not control severity of accident severe consequences risks over time

~~Displacement of -1 -1 -1 -1 1 -1 -1 -1 1 recreational fishing~~

Same risk to present Stakeholders generally have Area displaced tends to be As long as area of wind farm and future opportunities to influence Risks stay the relatively small in comparison to is not prime or irreplaceable generations, lifespan fishers location and size of wind farm; WF would have inequitable but same or decrease the much larger extent of fishing Not fatal fishing grounds, impact can Local not global impact of wind turbines is 20- generally don't they tend to have some control in small impact on rec fishermen as people adjust grounds, this tends not to be not be reduced by moving fishing 30 years so minimal choose this relation to displacement and to change a dreaded concern effort elsewhere risk to people in consequently impact on fishing distant future

Displacement of -1 -1 -1 -1 1 -1 -1 -1 1 commercial fishing Area displaced is small relative Impact easily reduced by WF would have inequitable but fishers Same as above in relation to to size of bay, this is not a Not fatal moving commercial fishing small impact on commercial Local not global impact See above See above generally don't commercial fishing dreaded concern effort elsewhere fishermen choose this

~~Displacement of -1 -1 -1 -1 1 -1 -1 -1 1 recreational boating~~

No expressions of dread found in WF would have inequitable but Impact easily reduced by boaters Same as above in relation to literature in relation to small impact on recreational Not fatal recreational boating Local not global impact See above See above generally don't impact on fishing displacement of recreational boating; rec boaters likely elsewhere choose this boating wouldn't benefit much Negative impact on -1 -1 -1 1 1 -1 -1 -1 1 tourism

Results are inconclusive regarding if wind farms negatively impact No expressions of "dread" per se Not easily reduced: tourism tourism. It is a common concern, found in literature in relation to operations would likely need but tour operators control what negative impact on tourism. tour operators to change their operations Concerns raised about impacts of they advertise and show so they People are concerned, but we Not fatal Local not global impact See above See above likely do not that currently focus on WF on tourism could capitalize on the green tech did not find documentation of choose this wildness of land and aspect of farm. Many tourists widespread anxiety or fear (aka seascape may want tours of the farm dread). (Lilley, 2010).

Negative visual 1 -1 -1 1 1 -1 -1 -1 1 impact Placing the turbines further Dread or fear does not offshore to reduce visual The negative affective reaction to characterize most people's impact is not feasible with People living, working and

Potential Ecosystem Service Consequence visual impact is subjective so not attitudes to a WF. Many dislike Not fatal existing technology given recreating closer to coast would Local not global impact See above See above XXXX controllable and don't want it but it's not a water depths at distances at experience greater visual impact source of dread which farm would not be visible from land Impact to seabirds 1 1 1 1 1 -1 -1 -1 1 Extensive studies on bird migrations have been Birds may avoid People strongly value region's conducted to inform siting of area around People tend not to control bird high density of nesting sea birds, Some bird mortalities are WFs pose a higher risk to birds WFs. Once constructed, few turbines not voluntary behavior. Perception of high they are highly concerned with associated with wind than other marine species. No Local not global impact See above options currently exist to (Lindeboom, 2011) exposure likelihood of collisions development that could harm turbine collisions benefit to birds. reduce risk of bird collisions and learn to fly birds populations with commercial scale below or above modern turbines Impact on marine 1 1 1 1 1 -1 -1 -1 1 mammals

Concerns raised about impact to People dread potential harm to Perception of fatal marine mammals in interviews WF construction whales as evidenced by strong collisions (although none and past studies. Ecological phase has most Can not control marine mammal affective response in interviews have been documented in Interviewees do not know of studies suggest marine mammals acute impacts, behavior with regards to wind and to whale strandings and WF studies); perception not voluntary technologies to safely keep may benefit from increased food Local not global impact See above operations have turbines, collision is a common deployment of volunteer time that electromagnetic fields exposure whales away from turbines availability wind turbines. A small minimal impact concern and resources to reduce from underwater cables minority of interviewees (Snyder and fatalities of common whale could effect whale wondered if wind farm could Kaiser, 2009) strandings in bay strandings decrease whale strandings)

~~193~~

Factor 2 Risk factor Are the consequences Do the people exposed to the Are the consequences of Is the hazardous consequence Is the risk known to science? question observable? consequences know about it? exposure delayed? new to science? Diminishes risk Risk known to science (-) Observable (-) Known to those exposed (-) Effect immediate (-) Old Risk (-) perception Example Cars, bicycle Flooding Smoking Flooding Car, bicycle Unknown to those exposed Increases risk Risk Unknown to science (+) Not observable (+) Effect delayed (+) New Risk (+) perception (+) Fracking: impact on underground ecosystems Example Long term impact of fracking Radon (at least initially) Exposure to many pollutants Fracking and water is hard to observe

~~Displacement of -1 -1 -1 -1 -1 recreational fishing~~

Impact of displaced recreational Given visibly of WFs, this impact past developments have fishing has been studied, easy displacement can be would be known to those No time delay displaced fishing effort, e.g., for people to imagine it has observed displaced aquaculture, shipping known consequences

Displacement of -1 -1 -1 -1 -1 commercial fishing Given visibly of WFs, this impact past developments have Impact of displaced commercial displacement can be would be known to those displaced fishing effort, e.g., fishing has been studied observed displaced aquaculture, shipping Displacement of -1 -1 -1 -1 -1 recreational boating Given visibly of WFs, this impact past developments have Impact of displaced recreational displacement can be would be known to those No time delay displaced boating, e.g., fishing has been studied observed displaced aquaculture, shipping Negative impact on -1 -1 -1 -1 -1 tourism Tourism impacts have been studied. Studies suggests no or Given visibly of WFs, this impact impact on tourism can be past developments have minimal impact to tourism. would be known to those No time delay observed impacted tourism Results are not conclusive across displaced all study locations. Negative visual -1 -1 -1 -1 -1 impact surveys can be used to Given visibly of WFs, this impact Visual impact from offshore assess attitudes towards would be known to those No time delay visual impact not new to science wind farms has been studied Potential Ecosystem Service Consequence visual impact displaced Impact to seabirds 1 -1 -1 -1 -1 scientists have identified mechanisms underpinning impacts to seabirds are This impact could be measured impact to seabirds not new to impacts to seabird in other No time delay observable and known science locations from offshore wind farms Impact on marine 1 -1 -1 -1 -1 mammals scientists have identified mechanisms underpinning impacts to marine This impact could be measured impact to marine mammals not impacts to marine mammals in No time delay mammals are observable and known new to science other locations from offshore wind farms

~~194~~

~~See: https://youtu.be/w_JYLRHi_Bc~~

~~195~~

~~Appendix E Choice experiment consent form~~

Principal Investigator Co-Investigator Dr. Kai Chan Sarah Klain, PhD Candidate University of British Columbia University of British Columbia 2202 Main Mall 2202 Main Mall Vancouver BC, Canada Vancouver BC, Canada

~~We are conducting a survey about people’s preferences based on different text and image- based descriptions. The survey will take approximately 20 minutes.~~

~~This research will contribute towards Sarah Klain’s PhD dissertation.~~

~~Sponsor This research project was made possible by a grant from the Social Science and Humanities Research Council of Canada (SSHRC).~~

~~Purpose You are invited to take part in this research because you are a resident of New England and we are interested in New Englanders preferences and opinions.~~

~~Study Procedures If you consent, you will be directed to a survey and you will make choices based on your personal preferences. We will also ask a few demographic and attitude-related questions.~~

Potential Risks To minimize and avoid psychological stress, the confidentiality of the information that you share is guaranteed and you are free to stop participating in this survey at any point. We ask for your m-turk worker id, but no information that reveals your identity.

Potential Benefits Information from your participation in this study may inform policy and development options. You may find the survey educational. If you are interested in receiving a digital copy of the output of this research, please email Sarah Klain at XXX.

~~Confidentiality We are not collecting information that could identify who you are. The M-Turk system protects the anonymity of its workers.~~

~~196~~

~~Remuneration/Compensation You will be paid $1 to complete this survey via the M-Turk system.~~

~~Contact for information about the study If you have questions or want to know more information about this study, please email Sarah Klain XXX or Kai Chan at XX.~~

Contact for concerns about the rights of research subjects If you have any concerns or complaints about your rights as a research participant and/or your experiences while participating in this study, contact the Research Participant Complaint Line in the UBC Office of Research Services at 604-822-8598 or if long distance e-mail [email protected] or call toll free 1-877-822-8598 (Toll Free: 1-877-822-8598).

~~Consent Your participation in this study is entirely voluntary and you may refuse to participate or withdraw from the study at any time without jeopardy to your employment.~~

~~Clicking “I consent to participating in this study” indicates your consent in choosing to take this survey.~~

~~197~~

~~Appendix F Choice experiment Mechanical Turk request description~~

~~Requester: Sarah Klain~~

~~Qualifications Required: HIT approval rate (%) is higher than 50; Location is ME, MA, CT, NH, RI~~

~~Reward: $1.00 per HIT HITs available: 1~~

~~University of British Columbia~~

~~We are conducting a survey about people’s preferences based on different text and image- based descriptions. The survey will take approximately 20 minutes.~~

~~Make sure you know your M-Turk Id.~~

~~Responses will be checked before approval. Once approved, you will be paid $1.~~

~~Please follow these steps to complete the survey:~~

1. Accept the HIT 2. Open the survey in a different Tab or Window (right-click on link and select option): https://ubc.qualtrics[xxx] 3. Complete the survey. A Completion Code will be shown when you finish this survey. This code is necessary to process payment 4. Insert the Completion Code below:

~~Thank you for your interest!~~

~~198~~

~~Appendix G Choice experiment survey~~

~~Options for Electrifying the Future~~

~~Introduction~~

A wind farm is a cluster of wind turbines used to generate electricity. Based on US Department of Energy studies, coastal New England has strong and abundant offshore wind resources as shown in the map below.

~~Wind resource potential Poor Fair Rhode Good $ Island Excellent Outstanding~~

~~1. Have you seen a wind turbine in operation? o Yes o No~~

~~2. What is your attitude toward developing wind power in the US? o Very positive o Positive o Neutral o Negative 199~~

~~o Very Negative~~

~~3. In your opinion, construction of offshore wind turbines off the coast of your state should be: o Encouraged o Tolerated o Discouraged o Prohibited o Not sure~~

~~4. Would the presence of a visible offshore wind farm make you more or less likely to go to the coast for recreational purposes (e.g., beach-going, boating, fishing, or walking along the coast)?~~

~~o Much less likely o Less likely o No Difference o More Likely o Much more Likely~~

~~Choices of Electricity Sources~~

Research on how people make decisions shows that how people feel, their prior knowledge and their past experiences affect how they make decisions. We need to know if you take the time to read directions, otherwise the information you provide in this survey will not be useful. To demonstrate that you have read the instructions, for the next question on how you feel about wind turbines, please select “None of the above” as your answer.

~~Please check all the words that describe your feelings towards wind turbines:~~

~~Supportive Opposed Disinterested Skeptical Interested Afraid Concerned Curious Apathetic Enthusiastic Appreciative None of the Above~~

For the purpose of this survey, please assume that your state has committed to increase energy generation by 10%. Imagine that you have the opportunity to vote on either 1) An offshore wind farm with 100 wind turbines; or 2) A new coal or natural gas plant

~~200~~

Imagine that a wind farm is being considered for a site off the coast of your state. As part of the negotiation with various stakeholders, you and other residents are given shares worth $100 in the wind farm company (or cooperative) if the wind farm is developed.

~~A Google Earth visualization of an offshore wind farm. The eye altitude is 3 feet above the ocean. Typical offshore wind farm towers rise to around 360 feet above sea level.~~

If an offshore wind farm is built, assume a renewable energy fee would be added each month to your electricity bill. This fee would be used to offset construction and maintenance costs for the lifespan of the wind farm, which is about 25 years.

We will ask you to vote for your preferred option while assuming that: • The electricity generation option that receives the most votes will be constructed • Each energy option generates an equal number of job opportunities • Potential wind farm sites have equal wind resources • Wind farm locations are outside of bird migration pathways and distant from bird nesting areas • Engineers and biologists can create underwater structures as part of the tower, which supports the turbine blades. This tower could provide different levels of underwater habitat quality.

~~This wind energy company could be a:~~

~~201~~

• Cooperative: members own the business, all profits after taxes are given back to members • Private company or corporation: owned by share holders who appoint a board of directors who supervise the business • Municipal owned and operated initiative: the wind farm is publicly owned by the municipal government • State owned and operated initiative: the wind farm is publicly owned by the state government

~~202~~

~~Please consider the following set of options.~~

Op#on&A&& Op#on&B& Op#on&C& Wind&Farm! Wind&farm! Coal&or&Gas&Plant& No&Wind&Farm! Eﬀect&on& • Large!loss! • Large!gain! • No!wind!farm! marine&life! • 60%!decline!in!diversity!and! • 60%!increase!in!diversity! • Expansion!of!coal!or! abundance! and!abundance! natural!gas! • Turbine!structures!provide! • Turbine!structures!provide! • No!direct!impact!on! poor&habitat&for!underwater! excellent&habitat!for! marine!ecosystems! plants!and!animals,!e.g.,!an=> underwater!plants!and! • Associated!CO2!emissions! fouling!paint!used!on!tower! animals! contribute!to!ocean! ! ! acidiﬁca=on!!

~~Wind&farm& Private! Municipality!owned! Ownership!not!speciﬁed! Ownership! &! &!~~

~~Visibility& Highly!visible![play!movie]! Barely!visible![play!movie]! Built!on!land! from&shore! 1!mile!from!shore! ≥10!miles!from!shore!~~

~~Addi#on&to& $1! $20! $0! monthly& electricity& u#lity&bill! &! Which option would you vote for? I would vote for: 203~~

~~o Option A o Option B o Option C~~

~~[Repeat for a total of 8 Choice Sets. Each choice set varies the levels and attributes according to my orthogonal array]~~

Imagine that a wind project off your state’s coast was the first of numerous North American offshore wind projects. Would this influence your attitude towards the wind project? For example, suppose that building 300 wind farms off the coast from Connecticut to Maine could supply 30% of the electricity for New England coastal states. Together, these wind farms would have a substantially larger impact on the ocean than one wind farm. However, 300 wind farms could greatly reduce air pollution, foreign oil dependence, and reliance on fossil fuel linked to climate change and sea level rise. If you knew that the farm near your state’s coast was the first of many offshore wind farms, would you be more or less likely to support the wind farm?

~~1 2 3 4 5 |------|------|------|------| Less likely No effect on More likely to support my decision to support~~

~~Details about yourself to help us interpret our survey results~~

~~Are you female or male? o Female o Male~~

~~How old are you?~~

~~What is your zip code?~~

What is your race or ethnic origin? Check all that apply. o American Indian or Alaska Native o Asian o Black or African American o Hispanic, Latino or Spanish o Native Hawaiian or Other Pacific Islander o White European o Middle Eastern 204

~~o North African o Other ______~~

What is the highest level of education that you have completed? Please check one. o Grade school o Some high school o High school graduate o Some college credit o Associate degree o Bachelor’s degree o Graduate degree or Professional degree

Which category best describes your household income before taxes in 2014? o Less than $10,000 o $10,000-$14,999 o $15,000-$24,999 o $25,000-$34,999 o $35,000-$49,999 o $50,000-$74,999 o $75,000-$99,999 o $100,000-$124,999 o $125,000-$149,000 o $150,000-$174,999 o $175,000-$199,999 o $250,000 and above

~~What is your employment status? o Employed for wages o Self-employed o Out of work o A homemaker o Student o Retired~~

~~Have you heard of Mechanical Turk? To confirm that you are carefully reading instructions, please select: Yes, I am an MT worker.~~

~~o Never heard of it o No, what is that? o Vaguely, but I’m not sure o Yes, I am an MT worker o Yes, I have done many HITs~~

~~205~~

~~Please indicate your political affiliation: o Democratic party o Republican party o Independent o Other (please specify) o None~~

Do you recreate on the coast? This could be a range of coastal or ocean-based activities such as going to the beach, surfing, fishing, and/or boating. o Frequently, 20+ times/year o Sometimes, 10-20 times/year o Every now and then, 5-10 times/year o Rarely, 1-5 times/year o Never

~~Attitudes~~

~~On a scale of 1 (Strongly Disagree) to 5 (Strongly Agree), to what extent do you agree with the following statements?~~

~~Humans are severely abusing the environment. 1 ------2------3------4------5 Strongly disagree Disagree Unsure Agree Strongly agree~~

~~The balance of nature is strong enough to cope with the impacts of modern industrial nations. 1 ------2------3------4------5 Strongly disagree Disagree Unsure Agree Strongly agree~~

~~The so-called “ecological crisis” facing human kind has been greatly exaggerated. 1 ------2------3------4------5 Strongly disagree Disagree Unsure Agree Strongly agree~~

~~The earth is like a spaceship with very limited room and resources. 1 ------2------3------4------5 Strongly disagree Disagree Unsure Agree Strongly agree~~

~~If things continue on their present course, we will soon experience a major ecological catastrophe. 1 ------2------3------4------5 Strongly disagree Disagree Unsure Agree Strongly agree~~

~~206~~

~~On a scale of 1 (Strongly Disagree) to 5 (Strongly Agree), to what extent do you agree with the following statements?~~

Plants and animals, as part of the interdependent web of life, are like ‘kin’ or family to me, so how we treat them matters. 1 ------2------3------4------5 Strongly disagree Disagree Neutral Agree Strongly agree

Humans have a responsibility to account for our own impacts to the environment because they can harm other people. 1 ------2------3------4------5 Strongly disagree Disagree Neutral Agree Strongly agree

I have strong feelings about nature (including all plants, animals, the land, etc.); these views are part of who I am and how I live my life. 1 ------2------3------4------5 Strongly disagree Disagree Neutral Agree Strongly agree

I often think of some wild places whose fate I care about and strive to protect, even though I may never see them myself. 1 ------2------3------4------5 Strongly disagree Disagree Neutral Agree Strongly agree

~~There are landscapes that say something about who we are as a community, a people. 1 ------2------3------4------5 Strongly disagree Disagree Neutral Agree Strongly agree~~

How I manage the land, both for plants and animals and for future people, reflects my sense of responsibility to and so stewardship of the land 1 ------2------3------4------5 Strongly disagree Disagree Neutral Agree Strongly agree

I think about the forest and all the plants and animals in it like: A family of which I am very much a part o Yes, this is very much like how I think about the forest o Yes, this is like how I think about the forest o This is somewhat like how I think about the forest o This is somewhat unlike how I think about the forest o No, this is very unlike how I think about the forest

~~207~~

Beings to which we owe responsible citizenship and care o Yes, this is very much like how I think about the forest o Yes, this is like how I think about the forest o This is somewhat like how I think about the forest o This is somewhat unlike how I think about the forest o No, this is very unlike how I think about the forest

Something that I identify with so strongly that it makes me, me o Yes, this is very much like how I think about the forest o Yes, this is like how I think about the forest o This is somewhat like how I think about the forest o This is somewhat unlike how I think about the forest o No, this is very unlike how I think about the forest

A world that we must care for so that any damage doesn’t also negatively effect humans who depend on it elsewhere o Yes, this is very much like how I think about the forest o Yes, this is like how I think about the forest o This is somewhat like how I think about the forest o This is somewhat unlike how I think about the forest o No, this is very unlike how I think about the forest

~~Thank-you for completing this survey! Your opinions are important.~~

~~Here is your code to insert in Mechanical Turk to receive your payment: [XXXXX]~~

~~If you want to learn more about this research project and why we asked you certain questions, click on Optional Debrief below.~~

~~Completion~~

Thank-you for completing this survey! Your opinions are important. Here is your code to insert in Mechanical Turk to receive your payment: [XXXXX] Let us know if you have any insights for improving this survey.

~~Optional Debrief~~

This research was designed to assess people’s preferences when it comes to making trade-offs related to renewable energy development. We will use the results to estimate public levels of support for a renewable energy technology that could be designed to increase the abundance

~~208~~

~~and diversity of marine ecosystems. We are also testing to see if the public prefers one type of wind company ownership model over others.~~

In Europe and China, wind farm developers are building offshore wind farms on an industrial scale. Offshore wind farms have not yet been built in North America. The higher construction and maintenance costs of offshore as compared to land-based wind farms can be largely offset by increased electricity generation since offshore wind tends to be stronger and steadier than onshore wind. Currently, offshore wind farms cost more per unit of electricity generated than most coal, natural gas or hydroelectric power stations, but operating a wind farm does not generate carbon emissions nor does it impact river ecosystems.

~~If you’re interested in learning more about the science of offshore wind farms, here are some sources of information:~~

~~The National Renewable Energy Laboratory, which is within the US department of Energy: http://www.nrel.gov/wind/offshore_wind.html~~

~~The Natural Resources Defense Council, Renewable Energy for America site on offshore renewables: http://www.nrdc.org/energy/renewables/offshore.asp~~

~~209~~

Appendix H Variables in choice experiment Variables used in discrete choice experiment regression models including description and means for survey respondents (n = 400). Variable Description Type of Data Mean ASC Alternative-specific constant choice A = 1; choice B = 1; 0.67 choice C = 0 big.loss Choice attribute is 60% decline in 1 = yes; 0 = no 0.18 diversity and abundance small.lossw Choice attribute is 30% decline in 1 = yes; 0 = no 0.17 diversity and abundance small.gain Choice attribute is 30% increase in 1 = yes; 0 = no 0.19 diversity and abundance big.gain Choice attribute is 60% increase in 1 = yes; 0 = no 0.13 diversity and abundance state state owned wind farm 1 = yes; 0 = no 0.18 municipal municipal owned wind farm 1 = yes; 0 = no 0.17 privatew privately owned wind farm 1 = yes; 0 = no 0.20 cooperative cooperative owned wind farm 1 = yes; 0 = no 0.12 mi1w wind farm 1 mile from shore 1 = yes; 0 = no 0.16 mi4 wind farm 4 miles from shore 1 = yes; 0 = no 0.17 mi8 wind farm 8 miles from shore 1 = yes; 0 = no 0.20 mi10 wind farm > 10 miles from shore 1 = yes; 0 = no 0.14 bill cost of OWF as addition to monthly $1; $5; $10; $20 5.38 utility bill white Respondent is white 1 = yes; 0 = no 0.83 female Respondent is female 1 = yes; 0 = no 0.59 age Age of respondent 18-69 32.38 univ_degr Respondent has a university degree 1 = university degree or 0.66 more, 0 = less than university degree income Household income before taxes 1 to 12; 1 = less than $10k; 5.36 12 = more than $250,000 wages Employed for wages 1 = yes; 0 = no 0.56 self.emp Self employed 1 = yes; 0 = no 0.11 coast_rec Sometimes or frequently recreates at coast 1 = yes; 0 = no 0.36 (10-20+ times/year) wBase case used in effects coding.

~~210~~

~~Appendix I Factor Analysis by population~~

~~Factor analysis results from tourist sample~~

~~Factor analysis results from farmer sample~~

~~211~~

~~bau_nep abuse_nep~~

~~spaceship_nepother_rel clean_inst 0.5 other_met health_rel comm_rel wild_reliden_relresp_met kin_rel~~

~~iden_metkin_met tech 0.0 Factor2~~

~~extract_ins loss_instr decade_morright~~

~~0.5 bal_nep −~~

~~crisis_nep~~

~~−0.2 0.0 0.2 0.4 0.6 0.8~~

~~Factor1~~

~~Factor analysis results from M-Turk sample~~

~~212~~

~~Appendix J Scree plot~~

Scree plot including responses to five NEP statements and six relational value statements across all three populations. Parallel analysis, optimal coordinates and acceleration factors are different methods to determine the number of factors to retain (Ledesma, 2011).

~~213~~

~~Appendix K Graphical PCA results~~

~~Graphical PCA results using data on responses to relational value and NEP statement~~

~~214~~

~~Appendix L M-Turk Cronbach’s alphas~~

~~0.8~~

~~0.6~~

~~0.4 Cronbach's alpha 0.2~~

~~0.0~~

~~NEP (5) intrinsic (2) relational (6) metaphor (4) instrumental (3) Type of Environmental Value Prompt~~

Cronbach alphas for M-Turk population. Note the different number of prompts in each category as shown in parentheses after each environmental value type. We suggest testing additional intrinsic and instrumental value prompts.

~~215~~

~~Appendix M Variables on wind farm attitudes and indices of environmental value~~

Variable Description Likert Scale Descriptor Score att_w_US What is your attitude toward developing wind power in the U.S.? Very negative 1 Negative 2 Neutral 3 Positive 4 Very positive 5 oper Have you seen a wind turbine in operation? No 1 Yes 2 const_st In your opinion, construction of offshore wind turbines off the coast Prohibited 1 of your state should be: Discouraged 2 Tolerated 3 Encouraged 4 wf_rec Would the presence of a visible offshore wind farm make you more or Much less likely 1 less likely to go to the coast for recreational purposes (e.g., beach- Less likely 2 going, boating, fishing, or walking along the coast)? No difference 3 More likely 4 Much more likely 5 coast_rec Do you recreate on the coast? This could be a range of coastal or Never 1 ocean-based activities such as going to the beach, surfing, fishing, Rarely, 1-5 times/year 2 and/or boating. Every now and then, 5-10 3 Sometimes, 10-20 times/year 4 Frequently, 20+ times/year 5 first_st Much less likely to support 1 Imagine that a wind project off your state’s coast was the first of numerous North American offshore wind projects. Would this influence your attitude Less likely to support 2 towards the wind project? For example, suppose that building 200 offshore wind No effect on my attitude 3 farms could supply 30% of the electricity for New England coastal states. More likely to support 4 Together, these wind farms would have a substantially larger impact on how Much more likely to support 5 people currently use the ocean and the ocean environment than one wind farm. However, 200 wind farms could reduce air pollution and reliance on fossil fuels linked to climate change and sea level rise. If you knew that the farm near your state’s coast was the first of many offshore wind farms, would you be more or less likely to support the wind farm? mean_nep Mean response to New Environmental Paradigm prompts Strongly Disagree 1 Disagree 2 Neither Agree nor Disagree 3 Agree 4 Strongly Agree 5 mean_rel Mean response to relational value prompts No, this is very unlike how I think about the ocean 1 This is somewhat unlike how I think about the ocean 2 This is somewhat like how I think about the ocean 3 Yes, this is like how I think about the ocean 4 Yes, this is very much like how I think about the ocean 5 mean_inst Mean response to instrumental value prompts Strongly Disagree 1 Disagree 2 Neither Agree nor Disagree 3 Agree 4 Strongly Agree 5 mean_met Mean response to metaphor value prompts Strongly Disagree 1 Disagree 2 Neither Agree nor Disagree 3 Agree 4 Strongly Agree 5 mean_mor Mean response to moral/intrinsic value prompts Strongly Disagree 1 Disagree 2 Neither Agree nor Disagree 3 Agree 4 Strongly Agree 5

~~216 From M-Turk Sample~~

~~Appendix N Wind farm attitudes~~

~~If you knew that the farm near the coast of your state was the first of many offshore wind farms, would you be more or less likely to support the wind farm? 40%~~

~~30%~~

~~20%~~

~~10%~~

~~Less likely to support More likely to support No effect on my attitude Much less likely to support Much more likely to support~~

~~Have you seen a wind turbine in operation? 0.8~~

~~0.6~~

~~0.4 Percent~~

~~0.2~~

~~0.0~~

~~No Yes Response~~

~~217~~

~~Appendix O Distribution of responses to value prompts To what extent do you agree with these statements?~~

~~Space Extract Loss Clean Kin Resp Inden Other Decade Rigth Abuse Bal Crisis ship Bau Comm Wild Resp Iden Kin Health Other 5~~

~~4 F a r 3 mer 2 1~~

~~5 Instrumental~~

~~4 M− Metaphorical T~~

~~3 u Intrinsic/moral r k 2 NEP Response 1 Relational 1 = Strongly Disagree; 5 4 = Agree; 5 Strongly Agree~~

~~4 T ou r~~

~~3 ist~~

~~2 = Disagree; 3 Neither Agree nor 2 1~~

0100200200 1000 200200 1000 200200 01000 200200 1000 200200 01000 200200 01000 200200 01000 200200 1000 200200 01000 200200 01000 200200 01000 200200 1000 200200 01000 200200 01000 200200 01000 200200 1000 200200 01000 200200 01000 200200 01000 200200 1000 200200 count

~~218~~

~~Appendix P Detailed site descriptions~~

~~Case 1. Block Island: The Ocean State’s Offshore Wind Farm Pioneers~~

Construction began on Deepwater Wind’s 30 MW, five-turbine wind farm three miles off the coast of Block Island in the summer of 2015 after a relatively smooth project development process compared to the nearby Cape Wind proposal. This can be attributed to many factors, including the groundwork established by the Rhode Island Coastal Resources Management Council’s Rhode Island Ocean Special Area Management Plan (SAMP) shortly before the project was proposed (Nutters and Pinto da Silva, 2012). Also, the relatively small scale of the Block Island project likely contributed to its ability to move forward first. The Block Island Wind Farm consists of five turbines compared to Cape Wind’s 130, the anticipated economic impact on electric rates is smaller than Cape Wind’s, and it is a multi-million dollar project while Cape Wind is a multi-billion dollar project (Smith et al., 2015). The Block Island Wind Farm also benefited from the state’s long-term contracting legislation, as well as minimal federal regulatory review due to the project’s location within state waters. While not without its opponents (McGlinchey, 2013), this project has been met with support from island leaders, a local Indian tribe, environmentalists and fishermen, in part due to well-defined benefits (Economist, 2015).

Timing played a key role in the success of this project. Creating and disseminating the SAMP before the wind farm was proposed meant that information about state waters was already readily available and accessible and had been discussed with key stakeholders (Nutters and Pinto da Silva, 2012), including the town council of New Shoreham on Block Island, which actively followed and contributed to the SAMP process. When Deepwater Wind proposed a wind farm in Rhode Island’s state waters, the New Shoreham Town Council was tasked with reviewing the proposal and representing the community’s interests and concerns. The town council recognized that it did not have energy experts on staff to review the associated technical documents within the structure of the regulatory process. To prevent a defensive David versus Goliath mentality 219

(i.e., the small island community standing up to a large, well-financed development corporation), Deepwater Wind and the town council discussed the town’s need for additional technical capacity to make the proposed project more accessible and understandable to residents. The town selected and hired consultants to represent their interests and Deepwater agreed to reimburse the town for the expense of these consultants (Island Institute, 2012a).

These consultants served the function of a bridging organization between the developers and the island community members. The consultants translated pertinent technical details and locally relevant information to the town council. They shared information with the broader community, fielded questions at community meetings, listened to community concerns and translated these concerns into comments during the formal regulatory processes. The expertise of the consultants provided the town council with greater confidence that community concerns would be better integrated into the wind farm planning processes. A New Shoreham Town Council Member recognized the importance of readily available information, hiring a trusted communicator and securing community benefits:

The community [of Block Island] benefited greatly from the sharing of information via the Ocean SAMP process, and by Deepwater Wind's commitment to putting in place a trusted liaison as conduit for information... By employing [the liaison] and locating his office on Block Island, Deepwater Wind was able to provide "up to the minute" information and build relationships of trust. This was critical to success. By negotiating with the developer a number of key community benefit items, the Town of New Shoreham became a partner (albeit small) in the project, not just a passive venue to be utilized [or] exploited…We became educated, conversant, increasingly confident, and responsible citizens as we faced each phase of the process… We learned that even a small island community can lead by example… There is no end to what needs to be learned and stewarded.

~~Local stakeholders, government officials and Island Institute staff were convinced that locally- relevant community benefits played an important role in the success of this project. Once the~~

~~220~~

farm is built, Block Island will, for the first time, be connected to the mainland grid. Deepwater Wind anticipates that this wind farm and the submarine transmission cables connecting the turbines and the island to the mainland electricity grid will lower the island’s electricity costs by 40% (Economist, 2015), which was a driver in garnering local support for the project.2 The project developer, Deepwater Wind, anticipates that this wind farm and the submarine transmission cables connecting the turbines and the island to the mainland electricity grid will reduce the island’s electricity costs (Smith et al., 2015). As a result, once the wind farm is completed, Block Island will no longer need to transport and burn approximately one million gallons of diesel fuel to power the island’s generators (Economist, 2015). The town negotiated to have fiber optic strands included in the electricity cable bundle that were provided for the town. Faster Internet service will benefit residents and businesses that have struggled with the slower microwave-based broadband, particularly during the busy summer months. Deepwater Wind and New Shoreham have also developed a formal Community Benefit Agreement (CBA) in which the wind farm company will pay for improvements to town infrastructure where the cable comes ashore. Further, the project is expected to generate 300 jobs during the construction phase, including opportunities for local mariners and fishermen (Smith et al., 2015).

~~Case 2. Martha’s Vineyard: Moving forward with a Cooperative Approach~~

Vineyard Power was an outgrowth of Martha’s Vineyard’s Island Plan, a sustainability strategy that the Martha’s Vineyard Commission completed based on input from thousands of island residents in 2009 to “create the future we want rather than settle for the future we get” (MVC, 2009, p. 1). Eight years after the controversial Cape Wind offshore wind project had been proposed, the plan included a recommendation to create a community-owned renewable energy cooperative so islanders could have more autonomy over their energy production and better ensure community benefits associated with renewable energy development. To date, Vineyard Power has developed five commercial-scale solar photovoltaic projects on Martha’s Vineyard

2 This anticipated cost reduction estimate did not account for the 2014 dip in oil prices. The offshore wind farm, however, is anticipated to reduce the volatility of electricity prices on the island. In the long term, natural gas and oil prices are expected to rise (EIA, 2015). 221

~~and continues to look to multiple renewable energy technologies going forward, including offshore wind.~~

In 2009, Vineyard Power began recruiting members. The price of a membership in the coop escalates over time, beginning at $50 and currently at $200 in 2015. People joined for social benefits (e.g., inclusion in the decision making processes in an island-owned, action-oriented group to create a more sustainable energy future for their community) and financial rewards (e.g., ownership and control of local renewable energy projects and stabilized electricity prices once a large-scale renewable energy project is developed) (Nevin, 2010). The cooperative’s community benefits are embedded in the cooperative’s mission: “to produce electricity from local, renewable resources while advocating for and keeping the benefits within our island community” and the organization’s vision “to be Martha's Vineyard's community-owned energy cooperative” (VPC, 2015).

Vineyard Power members have made community benefits a central theme in the development of this offshore wind farm. Lack of perceived community benefits, arguably, played a more minor role in Cape Wind, an earlier Massachusetts-based offshore wind farm proposal that has stalled due to lawsuits, regulatory issues and problems with its Power Purchase Agreement (PPA). Learning from the Cape Wind experience, Vineyard Power initially developed a wind farm ownership model influenced by the project design and financing structure of the community- owned Fox Islands Wind Project on Vinalhaven Island, Maine where the size of the project was linked to the amount of power consumed by the island (personal communication Peckar, 2015b). The complexity, scale and scope of the currently proposed offshore wind farm, which could be as large as 2,000 MW (Smith et al., 2015), vastly exceeds the three-turbine Fox Islands Wind Project yet the focus on local control and benefit remains.

In January 2015, BOEM auctioned the rights to lease offshore wind in areas in federal waters south of Martha’s Vineyard. Offshore MW received a 10% discount on their bid price because they had executed a Community Benefit Agreement (CBA) with Vineyard Power. The CBA outlined opportunities to investigate local benefits to the island including job creation, an

~~222~~

~~operations and maintenance facility, and local equity ownership in the project (VPCOMW, 2015).~~

The President of Vineyard Power Cooperative reinforced the importance of community engagement, providing accessible information and community benefits when he said: “Vineyard Power has always advocated for an open, community-based approach in the development of renewable energy projects. We have been an extremely active participant throughout the BOEM offshore wind leasing process and provide updates and information to local municipalities, businesses and residents of our island to ensure our community and stakeholders remain engaged. We also believe that any offshore wind farm development in our surrounding waters should provide local benefits. We took control of our energy future and decided to be an active participant in the process. Through years of outreach with our members, local legislators and the local municipalities, BOEM recognized the nation’s first Community Benefit Agreement between our organization and Offshore MW. Through this CBA, we will ensure that our island community’s local economy will remain strong through local ownership, and job creation.”

In earlier stages of the project’s development, the cooperative hosted an interactive offshore wind map viewer on its website to not only inform but also solicit preferences from coop members and other engaged island residents to find a suitable location for the wind farm. This website provided readily available and appropriate information while encouraging participation in sharing local values related to proposed locations. The website provided information about visual, ecological and human use impacts based on various proposed sites, including data collected from local sources such as island fishermen. The cooperative also hosted a series of community meetings to share wind farm visualizations and solicit feedback (Peckar, 2015a).

Case 3: Monhegan Island: Confronting deep water challenges The tumultuous path of offshore wind in Maine provides insights regarding mutual learning, timing and accessibility of information. In 2009, Maine set ambitious goals to become a national leader in ocean energy (MCP, 2009) and created opportunities for the development of offshore

~~223~~

wind and tidal energy demonstration projects in both state and federal waters (MPUC, 2010). In each of these jurisdictions, discussions of offshore wind had implications for the island of Monhegan, a remote community 12 miles out to sea with a year-round population of about 60 and some of the highest energy costs in the nation at ~$0.70 kWh as compared to ~$0.15 kWh for mainland residential electricity in Maine (MPUC, 2015).

In state waters, Maine took initial steps to engage stakeholders in its strategy to expedite the development of the industry by designating three research and demonstration test sites within state waters. Representatives of Governor Baldacci’s Ocean Energy Task Force worked with the Maine Coastal Program (MCP) within the Maine State Planning office to host a series of public meetings and “kitchen table” (i.e., small and informal) discussions along the Maine coast where sites were under consideration. They incorporated scientific data and local knowledge into their assessment process by making mutual learning accessible. For example, when MCP and other state agency staff traveled to Monhegan to gather feedback on the potential to create a site two miles from the island, they met with fishermen in a local fish house. They asked fishermen to rank their fishing activity effort around the island in order to identify a site of least impact for the turbines.

Efforts to site offshore wind in nearby federal waters underscored the importance of timing and availability of information. On September 1, 2010, the Maine Public Utilities Commission (PUC) began a 16-month process during which they solicited and reviewed bids for and public comments on a long-term power purchase agreement. This extended period of time provided an opportunity to engage stakeholders prior to the announcement of a developer and the location of a site. During this time, the Island Institute worked as a bridging organization to facilitate mutual learning through the Offshore Wind Energy Information Exchange, an outreach and education initiative to inform and engage coastal and marine stakeholders, developers, and decision-makers on the potential for offshore wind energy development in the Gulf of Maine. The initiative included deliberative learning experiences, such as exchange trips to fishing communities as well as a wind farm, the human use mapping project Mapping Working Waters (Island Institute, 2009), information sessions at the annual Fishermen’s Forum in Maine (Island Institute, 2012b)

~~224~~

and readily available and understandable fact sheets (Island Institute, 2012a). These efforts provided coastal stakeholders and industry representatives with a baseline understanding of community priorities as well as the offshore wind industry, while creating an opportunity for stakeholders to meet each other informally and build relationships.

In January 2013, Maine PUC announced its selection of an unsolicited proposal from Statoil – a multinational corporation specializing in offshore energy infrastructure – for testing floating turbine technology in federal waters in the state’s Midcoast region. By this time, marine users and other stakeholders in the area had already participated in education and information exchange opportunities, preparing them to more proactively and constructively engage in discussions with the developer and decision-makers (Island Institute, 2015).

Later in 2013, the University of Maine entered a federal funding competition with a new scope of activities at the Monhegan test site. Subsequently, the Maine Legislature directed the PUC to reopen the bidding process so that the University of Maine could submit a proposal on an accelerated timeline, and Statoil withdrew its proposal for a project in federal waters. While these developments had statewide implications, this impacted Monhegan by significantly limiting the timeframe in which the community could learn about the change in scope from small-scale portable to large-scale, semi-permanent turbines. The PUC opportunity, which prompted many islanders to learn of the change in project scale, was announced during the summer, which is the island’s busiest time of year.

The accelerated timeline and need for information initially strained relations between the island community and Maine Aqua Ventus (MAV), the University-led consortium developing the larger project, but both parties quickly committed to improve communications. The first step was to clarify points of contact and expectations for communications so that MAV could be certain that project updates were being shared widely. Island leaders created the Monhegan Energy Task Force (METF) as a way to prioritize information that the community needed and facilitate discussion of community benefits associated with the proposed offshore wind project. METF and MAV engaged in weekly phone calls to enhance the flow of information and worked to develop

~~225~~

an expectations document to ensure timely project communications. During this time, both parties looked to Block Island for examples of how information was shared and community benefits arranged. MAV also began to host semi-regular open house sessions on the island during which residents and visitors could have more extended discussions about aspects of the project. In late 2015, MAV received additional federal funding ($3.7 mill) to continue refining their floating turbine designs (Turkel, 2015).

The co-chair of the Monhegan Energy Task Force (METF) reflected on dispelling misconceptions and improving communication between islanders and wind farm developers: “As we try to keep our very small community running, it is easy to get lost in the “doing” and not the “talking.” While dealing with Maine Aqua Ventus, the greatest challenge we faced was how to quickly get correct information to the community. The key for Monhegan Energy Task Force was to develop a plan for sharing information and for making research resources accessible. We co-authored a communications MOU with Maine Aqua Ventus, developed a website, sent mailings, and created an email list of stakeholders – making it possible to “tell” while we were doing. Open communication between the community and Monhegan Energy Task Force paired with open communication between Monhegan Energy Task Force and Maine Aqua Ventus helped all parties keep up to date and kept misinformation to a minimum.”

Based on our interviews, some residents still have concerns about the Monhegan offshore wind project but the developer and community have laid a more solid foundation upon which future communication can take place.

~~226~~