Download/Isaaa-Brief-53-2017.Pdf

NARRATIVES OF CONFLICT IN AGRICULTURAL BIOTECHNOLOGY POLICY IN INDIA

JUHI HUDA

B.A., University of Pune, India, 2007

M.A., University of Pune, India, 2009

M.A., University of Nevada Reno, 2013

A thesis submitted to the

Faculty of the Graduate School of the

University of Colorado in partial fulfillment

of the requirement for the degree of

Doctor of Philosophy

Environmental Studies Program

2019

This dissertation entitled:

Narratives of Conflict in Agricultural Biotechnology Policy in India

written by Juhi Huda

has been approved for the Environmental Studies Program

Committee Chair:

______Dr. Deserai Anderson Crow, Ph.D.

Committee Members:

______Dr. Sharon Collinge, Ph.D.

______Dr. Peter Newton, Ph.D.

______Dr. Elizabeth A. Shanahan, Ph.D.

______Dr. Christopher M. Weible, Ph.D.

Date:

The final copy of this dissertation has been examined by the signatories, and we find that both the content and the form meet acceptable presentation standards of scholarly work in the above-mentioned discipline.

IRB protocol # 16-0414

ii ABSTRACT

Huda, Juhi (Ph.D., Environmental Studies)

Narratives of Conflict in Agricultural Biotechnology Policy in India

Thesis directed by Associate Professor Deserai Anderson Crow

The Narrative Policy Framework (NPF) focuses attention on narratives in policy debates and their empirical analysis. While NPF has become an important and accepted approach to studying the policy process, the majority of research applies it to policy and linguistic contexts of the United States, which limits its generalizability and responsiveness to cultural specificity. In this dissertation, I primarily endeavor to push the NPF forward by refining its concepts and testing its transportability by applying it to the policy subsystem of agricultural biotechnology policy in India. Secondly, I examine how the information contained in these narratives pertaining to policy problems, solutions, science, ethics, risk, and other factors is used strategically in the contentious issue of agricultural biotechnology. I use a mixed method approach drawing on quantitative and qualitative content analysis and qualitative interview data to empirically analyze how stakeholders use policy narratives to interact and advocate for policy change.

Using the case study of commercialization of a genetically modified crop, Bt eggplant, I examine media coverage from leading English newspapers in India to explore the strategic use of narrative variables. Findings indicate that policy narratives do not always contain a full suite of narrative components and yet may be among the most common messages received by the public and political actors emphasizing a need to further refine the definition of policy narratives and consider which narratives are important from empirical and audience reception perspectives. I examine the NPF assumption that narratives have generalizable narrative elements irrespective of variation in linguistic context and test its transportability. Findings lend support to its

iii transportability outside the English language and indicate variation in use of narrative elements across languages. Examining setting and plot, I explore further into the policy issue of agricultural biotechnology and focus on the evidence in support of claims about risks and benefits and explore moral notions of risk. Findings indicate that stakeholders use different sources of evidence and proponents de-emphasize risks and exclusively highlight benefits while opponents invoke multi-dimensional risk; and risk perceptions of stakeholders are influenced by moral notions of risk. In sum, the findings inform both the theoretical study of NPF and communication practices of stakeholders.

To kismet

v ACKNOWLEDGEMENTS

A number of people made this dissertation possible, contributed to my professional development, and provided invaluable support. I can never adequately thank all of them and I apologize for inadvertently leaving anyone out.

To my dissertation chair – Dr. Deserai Crow: Deserai, thank you for taking a chance on me. Without your unwavering support and constant guidance, I am certain I would not have made it this far. Your mentorship is invaluable, and I could not have asked for a more thoughtful, caring, driven, and organized advisor. Your meticulous planning and organization made the dissertation process smoother than I could have ever imagined and helped turn dissertation chapters into published journal articles. Thank you for the research opportunities, for your advocacy on my behalf, and for your words of advice and encouragement at each and every turn. Thank you!

To my committee members – Drs. Elizabeth Shanahan, Christopher Weible, Peter Newton, and Sharon Collinge: Liz, thank you for your generous feedback on each of my papers from this dissertation. Your thoughtful advice and your desire to advance the NPF forward pushed this dissertation further than I’d imagined it could go. Thank you for always asking the hard questions. Chris, thank you for your feedback on my papers. Liz and Chris, your thoughtful and detailed suggestions helped transform my ideas into publishable manuscripts. Pete, thank you for the insightful conversations on food and agriculture, for your enthusiasm about my project, and for your support. Sharon, thank you for teaching ENVS 5000 in Fall 2014. It was the module on genetic engineering that triggered my journey into agricultural biotechnology. Your enthusiasm for the topic encouraged me to explore it more.

To my research group – Elizabeth Koebele, Lydia Lawhon, John Berggren, Adrianne Kroepsch, and Rebecca Schild: Collaborating with you all helped me gain so much research and publication experience. Thank you!

To all those I met and who helped me in the field in India – Thank you to the interview participants who shared their valuable time and insight. Thank you to those at Dainik Jagran, especially Mr. Radhe Shyam and Mr. Vijay Singh, who made the Hindi media data collection possible. And thank you to friends, Nisha Garud, Ranjeeth Rane, Preeti Virkar, Hrishi Chandanpurkar, who leveraged contacts to help me connect with interview participants.

To friends, colleagues, faculty, and staff in the Environmental Studies Program (ENVS) and in the Program for Writing and Rhetoric (PWR) – Members of the graduate committee at ENVS, thank you for funding so much of my research. It would not have been possible without your generous support. Thank you also for the teaching opportunities. Dr. Max Boykoff, thank you for the opportunity to collaborate. Penny Bates and Jean Lindahl, thank you for making the administrative processes so much smoother. Thank you to PWR - Dr. Steve Lamos, thank you for your excellent guidance on pedagogical practices, thank you to Drs. Steve Lamos and Linda Nicita for the teaching opportunities, and to Melynda Slaughter for helping to navigate the administrative process at PWR.

vi To mentors – I wouldn’t have started on this path to my Ph.D. if Dr. Scott Slovic and Dr. Derek Kauneckis (then at University of Nevada Reno) had not guided me in pursuing my own passion and interests. Thank you for your mentorship and guidance. I am also grateful to Dr. Chandrani Chatterjee and (late) Dr. Aniket Jaaware for their early mentorship at the University of Pune. They triggered my journey halfway across the planet.

To family – Ammi and Abba, thank you for not holding me back. To my brother, Yusuf, and sister-in-law, Renita, thank you for your love and support even if you, at times, had no idea what I was doing with my life. To Nani, possibly the strongest woman I’ve known. Wish you’d lived to see your granddaughter on this day. I miss you everyday. To Badi Khala and Choti Khala, thank you for your strength, encouragement, and unconditional love. To Aai, Aba, and the Chandanpurkar family, thank you for your love and acceptance. To the two felines in my life, Pan aka Pantalaimon and BoCo aka BoulderColorado. You’ve brought me so much joy and peace. Thank you for all the cuddles.

And finally, thank you to Hrishi. Without you, I know I wouldn’t be here. Thank you for believing in me, even when I didn’t.

This research has been financially supported by the Graduate School, the Center to Advance Research and Teaching in the Social Sciences (CARTSS), and the Environmental Studies Program at University of Colorado Boulder.

vii

“Words are events, they do things, change things. They transform both speaker and hearer; they feed energy back and forth and amplify it. They feed understanding or emotion back and forth and amplify it.”

- Ursula K. Le Guin, Telling is Listening, The Wave in the Mind: Talks and Essays on the Writer, the Reader, and the Imagination (Shambhala Publications, 2004)

viii TABLE OF CONTENTS CHAPTER 1: INTRODUCTION ...... 1 STRUCTURE OF THE DISSERTATION ...... 6 REFERENCES ...... 7 CHAPTER 2: NARRATIVES OF CONFLICT IN AGRICULTURAL BIOTECHNOLOGY POLICY IN INDIA: THEORY, CASE STUDY, RESEARCH GOALS, AND METHODS ...... 8 THEORETICAL ORIENTATION ...... 9 RESEARCH DESIGN: CASE STUDY AND METHODOLOGY ...... 13 RESEARCH GOALS AND FINDINGS ...... 18 METHODS: DATA COLLECTION AND ANALYSIS ...... 20 REFERENCES ...... 23 CHAPTER 3: AN EXAMINATION OF POLICY NARRATIVES IN AGRICULTURAL BIOTECHNOLOGY POLICY IN INDIA ...... 27 NPF AND PUBLIC POLICY SCHOLARSHIP ...... 28 NPF AND POLICY NARRATIVES IN THE MEDIA ...... 30 BROADENING THE NPF: APPLICATIONS IN NON-U.S. AND OTHER POLICY CONTEXTS ...... 34 AGRICULTURAL BIOTECHNOLOGY POLICY—THE CASE OF BT EGGPLANT IN INDIA ...... 35 RESEARCH METHODS ...... 37 RESEARCH FINDINGS AND DISCUSSION ...... 40 CONCLUSION AND FUTURE DIRECTIONS ...... 52 REFERENCES ...... 56 CHAPTER 4: POLICY NARRATIVES ACROSS TWO LANGUAGES: A COMPARATIVE STUDY USING THE NARRATIVE POLICY FRAMEWORK ...... 60 NARRATIVE POLICY FRAMEWORK, NARRATIVE ELEMENTS, AND THE MEDIA ...... 63 WHY STUDY POLICY NARRATIVES ACROSS LANGUAGES? ...... 67 BACKGROUND INFORMATION: BT EGGPLANT COMMERCIALIZATION PROCESS ...... 70 METHODS ...... 72 RESEARCH QUESTIONS AND OPERATIONALIZATION ...... 76 EXPECTATIONS, RESEARCH FINDINGS, AND DISCUSSION ...... 77 CONCLUSION ...... 97 REFERENCES ...... 100 CHAPTER 5: SOURCES OF EVIDENCE FOR RISKS AND BENEFITS IN AGRICULTURAL BIOTECHNOLOGY POLICY IN INDIA: EXAMINING SETTING AND PLOT IN POLICY NARRATIVES ...... 106 NPF AND PUBLIC POLICY SCHOLARSHIP: ...... 108 SOURCES OF EVIDENCE AND PERCEPTION OF RISKS AND BENEFITS ...... 108 EVIDENCE FOR RISKS AND BENEFITS IN AGRICULTURAL BIOTECHNOLOGY ...... 113 AGRICULTURAL BIOTECHNOLOGY IN INDIA: EVIDENCE OF RISKS AND BENEFITS IN BT EGGPLANT ...... 116 METHODS ...... 119 FINDINGS AND DISCUSSION ...... 125 CONCLUSION, LIMITATIONS, AND FUTURE DIRECTIONS ...... 137 REFERENCES ...... 140 CHAPTER 6: CONCLUSION ...... 146 SUMMARY OF MAJOR FINDINGS ...... 146 A BRIEF REFLECTION ON THE PROCESS ...... 150 AREAS OF FUTURE RESEARCH AND CLOSING THOUGHTS ...... 151

ix REFERENCES ...... 154 BIBLIOGRAPHY ...... 155 APPENDICES ...... 166 MEDIA CODEBOOK ...... 166 INTERVIEW PROTOCOL ...... 178

x LIST OF TABLES

CHAPTER 3 Table 3. 1: Search Terms, Newspapers, and Article Counts ...... 38 Table 3. 2: Categories of Problem Definitions in Times of India and Hindustan Times ...... 42 Table 3. 3: Policy Problems and Presence of Evidence, Solutions, and Risk/Benefit Information in Times of India and Hindustan Times ...... 43 Table 3. 4: Problem Definitions Used by the Losing, Winning, and Incomplete Narratives ...... 44 Table 3. 5: Narrative Variables across Losing, Winning, and Incomplete Policy Narratives ...... 46 Table 3. 6: Presence of Character Types in Times of India and Hindustan Times ...... 48 Table 3. 7: Use of Characters across Losing, Winning, and Incomplete Policy Narratives ...... 50 Table 3. 8: Narrativity Index for Losing, Winning, and Incomplete narratives; and for Times of India and Hindustan Times ...... 52

CHAPTER 4 Table 4. 1. Newspapers, Search Terms, and Article Counts ...... 74 Table 4. 2. Intercoder Reliability Scores ...... 75 Table 4. 3. Narrative Elements Across Hindi and English Media ...... 78 Table 4. 4. Appearance of Policy Problems over Time across Hindi and English Media ...... 80 Table 4. 5. Appearance of Policy Solutions over Time across Hindi and English Media ...... 84 Table 4. 6. Focus of Policy Solutions over time ...... 86 Table 4. 7. Appearance of Characters over Time across Hindi and English Media ...... 88 Table 4. 8. Use of Evidence over Time across Hindi and English Media ...... 95 Table 4. 9. Types of Evidence provided over time ...... 96

CHAPTER 5 Table 5. 1. Expectations ...... 125 Table 5. 2. Evidence in Media Data (n=99) ...... 126 Table 5. 3. Interview Data – Sources of Evidence ...... 128 Table 5. 4. Risks and Benefits in Media Data (n=99) ...... 131 Table 5. 5. Interview Data: Perceptions of Risks ...... 132 Table 5. 6. Interview Data: Perceptions of Benefits ...... 133 Table 5. 7. Interview Data – Moral Notions of Risk ...... 136

xi LIST OF FIGURES

CHAPTER 3 Figure 3. 1: Coverage of Types of Narratives ...... 39 Figure 3. 2: Focus of Coverage over Time in Times of India and Hindustan Times ...... 42

CHAPTER 4 Figure 4. 1. Issue Coverage on Similar Events across Hindi (n=19) and English Media (n=52) . 75 Figure 4. 2. Focus of Problem Definitions over time ...... 82 Figure 4. 3. Character Appearance over Time ...... 89 Figure 4. 4. Heroes and Villains Across Hindi and English Media ...... 91 Figure 4. 5. Appearance of Heroes over Time Across Hindi and English Media ...... 93

xii CHAPTER 1: INTRODUCTION

The green revolution in the 1960s led to an increase in agricultural production globally but half a century later the United Nation’s Food and Agricultural Organization (FAO)’s 2015 report entitled “The State of Food Insecurity in the World” estimated that about 795 million people are chronically undernourished (UNFAO, 2015). With the projected growth in population, overall food demand is expected to increase by more than 50% and animal-based foods demand by nearly 70% by 2050 (Searchinger, Waite, Hanson, & Ranganathan, 2018).

With world population growing rapidly, the number of food insecure people is expected to increase. While the green revolution of the sixties was characterized by an increased use of fertilizers, pesticides, and intensive irrigation techniques that enabled high-yielding varieties of crops, it also led to widespread soil pollution and massive depletion of groundwater reserves

(Rodell, Velicogna, & Famiglietti, 2009). There has been an increase in calls for a second green revolution to increase agricultural production.

Governing bodies are seeking ways to meet the growing demand for food sustainably.

Various solutions include increasing agricultural productivity, shifting dietary patterns, increasing cropping efficiency. Although food policies are focused on a general goal of fulfilling basic human needs, some proposals such as agricultural biotechnology are entangled in politically polarizing and contentious debates. Genetically modified (GM) crops such as glyphosate-resistant crops and Bacillus thuringensis (Bt) crops protect the crops from herbicides and pests respectively. This may potentially result in increased yields and reduced use of pesticides (Kishore, Padgette, & Fraley, 1992; Qaim & Zilberman, 2003). However, agricultural biotechnology has received a mixed reception globally.

1 As per a 2017 report, the United States continues to lead in the adoption of GM crops with ten crops planted in 2017, while Canada planted six. The European Union continues to lead the Western Hemisphere in rejection of agricultural biotechnology with a few exceptions in

Spain and Portugal. Commercialization of GM crops in Africa (thirteen countries) and Latin

America (eleven countries) has increased. Leading biotech countries in Asia and the Pacific region include India, followed by Pakistan, China, Australia, the Philippines, Myanmar,

Vietnam, and Bangladesh contributing to 10% of the global biotech crops. Lack of government approval has blocked several GM crops such as Golden Rice, Bt eggplant, and drought tolerant and insect resistant maize in different regions. Bt eggplant, commercialized in Bangladesh is banned in neighboring India. Regulatory approval for drought tolerant and insect resistant maize, a public-private collaboration of four countries in sub-Saharan Africa remains stalled, while

Brazil approved a virus-resistant bean in 2011 but the technology has not reached farmers

(ISAAA, 2017).

Deployment of biotechnology has introduced risk-related concerns, which has often culminated in struggles amongst stakeholders to influence policy outcomes. In the politics of risk, new technologies create new challenges for states to assess risk to public health and safety.

All stakeholders claim to have science and ethics on their side. These and other factors play a role in the decision-making process, but questions remain about how different factors actually contribute to decision-making in the face of risks that are introduced by new technologies.

India is an excellent country in which to study the agricultural biotechnology policy process and the role of different factors in the decision-making process. The aforementioned

FAO report states that “The highest burden of hunger occurs in Southern Asia, where as many as

281 million people are undernourished in the region” (3). India is particularly vulnerable. After

2 the green revolution in the 1960s, India became largely self-sufficient in agricultural production

(Das & Bhardwaj, 2015). However, rapid growth in population, reduction in arable land due to industrialization, and reliance on monsoons are some of the reasons leading to reduced agricultural production. In recent years, there has been a shortage of essential items leading to imports of goods such as sugar, lentils, and edible oil (Narain & Kumar, 2015). Agricultural biotechnology has increasingly been publicized as an important means of bringing about a second green revolution (Restall, 2014). However, the matter remains contentious as those against the adoption of GM crops suggest that increased import of goods may not necessarily be a result of factors such as a shortage in agricultural production but other factors such as flawed economic policies (Ghosh, 2016).

India has a controversial history with agricultural biotechnology with the establishment of the Department of Biotechnology in 1986; adoption of a GM crop—Bt cotton—in the commercial sector in 2002; and an indefinite ban on the first GM food crop—Bt eggplant—in

2010. The case of Bt eggplant is particularly important since it was the first genetically modified food crop to be considered for commercialization in India and since it has the same gene as the already commercialized, Bt cotton, it was widely believed that it would get swift approval.

However, due to protests from several stakeholders, an indefinite moratorium was placed on its release in 2010 that remains in effect.

Although previous studies (Herring, 2015; 2006) have examined the issue of GM crops in

India, there are no definite findings on the variation in reception. Given the large mix of stakeholders, including national and state governments, scientists, farmers, non-profit organizations, and the corporate sector, understanding the on-going debate about the advantages and disadvantages of the adoption of GM crops becomes a complex matter.

3 This dissertation emerged initially with the goal of understanding the role of factors such as narratives, science, ethics, and risks and benefits in agricultural biotechnology policy in India and, in particular, investigating how information contained in narratives pertaining to policy problems, solutions, science, ethics, risk, and other factors is used strategically in a contentious policy issue. Since the primary contention of the Narrative Policy Framework (NPF) is that

“understanding the role of narratives is critical to understanding the policy process” (Shanahan et al., 2018, p. 173), NPF seemed like the appropriate lens to use for this project. But as I formulated the project further, I realized I not only had the opportunity to examine narratives in the contentious policy issue of agricultural biotechnology policy to gain a better understanding of their role in the policy process, but also an opportunity to push the NPF forward given the relatively fewer applications of this framework to policy subsystems outside the United States and no application of the NPF in a different linguistic context.

The exploration that follows became more deeply embedded in policy process scholarship than in literature on agricultural biotechnology policy. To some extent, the narratives themselves took the front seat in this exploration while the case became a vessel to expand the scope of the NPF to examine its generalizability and transportability in a different policy subsystem situated in a different national and linguistic context. The findings in the empirical chapters contribute to the theoretical development of the NPF and advance the theoretical and empirical role of narratives in the policy process. But at the same time, they shed light on the critical role of information in agricultural biotechnology policy (such as evidence, risks, and benefits discussed in Chapter 5) or the lack of information (such as incomplete narratives discussed in Chapter 3) or the language used to convey that information (as discussed in Chapter

4).

4 My journey that culminated in this dissertation project was long and winding. I was fascinated with how a narrative is capable of structuring our perception of the environment. I joined the graduate program in Literature and Environment at the University of Nevada Reno and focused on developing a comparative methodology to study narrative strategies used in environmental discourse. In the wake of growing environmental challenges, I wanted my research to work toward practical solutions. As I searched for possible options that would allow me to draw from my expertise in narrative scholarship but at the same time have practical applications, I stumbled upon the Narrative Policy Framework. For me, it was the perfect platform to apply my training in environmental humanities to environmental social sciences.

I came to the graduate program at University of Colorado Boulder to study the NPF further. I was just not yet sure in which policy area. Fortunately for me, my first semester included a core Environmental Studies course that took a comprehensive look at the environmental aspects of food production and procurement. It linked policy, science, and values in food production systems and considered scientific evidence alongside policy and ethics dimensions of food production issues. As the course progressed, so did my interest in the food and agricultural policy subsystem and, in particular, the contentious debate surrounding agricultural biotechnology. To date, there have been no definite findings on the possible causes behind the variations in reception of genetically modified (GM) crops among nations and actors

(Herring, 2006, 2015). These twofold experiences motivated this project. Throughout my dissertation process, I hoped to contribute something new and practically oriented both to the existing scholarship on the Narrative Policy Framework and on agricultural biotechnology policy.

5 Structure of the Dissertation This dissertation is structured according to a three-paper dissertation format (chapters 3-

5):

Chapter 2 outlines the theoretical framework of the study situating NPF within the field of public policy scholarship discussing its historical development and main features in context of this study. It includes an in-depth discussion of the case study followed by research goals and methods.

Chapter 3 refines the definition of a policy narrative and examines the role of incomplete narratives in agricultural biotechnology policy in India.

Chapter 4 tests the transportability of narrative elements outside of the English language through its application to a linguistic context not studied before. It undertakes a comparative study of policy narratives on agricultural biotechnology policy in English and Hindi media.

Chapter 5 refines our understanding of setting and plot in policy narratives through an examination of how evidence of risks and benefits is used by stakeholders in the agricultural biotechnology policy process.

Chapter 6 summarizes the empirical chapters, reflects on lessons learned, provides some future research plans, and closing thoughts.

Since each of the empirical chapters was composed as an independent manuscript, the foundational theory and methodological and case study information (provided in chapter 2) is repeated and expanded further as necessary for each chapter to stand independently when published.

6 References

Das, K. N., & Bhardwaj, M. (2015, February 21). Modi bets on GM crops for India’s second green revolution. Reuters Publication. Retrieved from http://www.reuters.com/article/2015/02/22/us-india-gmo-insight- idUSKBN0LQ00Z20150222

Ghosh, A. K. (2016, April 6). Do we really need GM mustard in India? DownToEarth.

Herring, R. (2015). State science, risk and agricultural biotechnology: Bt cotton to Bt Brinjal in India. Journal of Peasant Studies, 42(1), 159–186.

Herring, R. J. (2006). Why did “Operation Cremate Monsanto” fail? Science and class in India’s great terminator-technology hoax. Critical Asian Studies, 38(4), 467–493.

ISAAA. (2017). Global status of commercialized biotech/GM crops in 2017: biotech crop adoption surges as economic benefits accumulate in 22 years. Ithaca, NY. Retrieved from https://www.isaaa.org/resources/publications/briefs/53/download/isaaa-brief-53-2017.pdf

Kishore, G. M., Padgette, S. R., & Fraley, R. T. (1992). History of Herbicide-Tolerant Crops, Methods of Development and Current State of the Art: Emphasis on Glyphosate Tolerance. Weed Technology, 6(3), 626–634. https://doi.org/10.2307/3987224

Narain, Y., & Kumar, S. K. (2015, June 25). Time for a Second Green Revolution. The New Indian Express. Retrieved from http://www.newindianexpress.com/columns/Time-for-a- Second-Green-Revolution/2015/06/26/article2886092.ece

Qaim, M., & Zilberman, D. (2003). Yield effects of genetically modified crops in developing countries. Science, 299(5608), 900–902.

Restall, H. (2014, November 21). Growing a Second Green Revolution. The Wall Street Journal. Retrieved from http://www.wsj.com/articles/growing-a-second-green-revolution- 1416613158

Rodell, M., Velicogna, I., & Famiglietti, J. S. (2009). Satellite-based estimates of groundwater depletion in India. Nature, 460(7258), 999–1002.

Searchinger, T., Waite, R., Hanson, C., & Ranganathan, J. (2018). CREATING A SUSTAINABLE FOOD FUTURE: A Menu of Solutions to Feed Nearly 10 Billion People by 2050.

Shanahan, E. A., Jones, M. D., McBeth, M. K., & Radaelli, C. M. (2018). The Narrative Policy Framework. In C. M. Weible & P. A. Sabatier (Eds.), Theories of the Policy Process (4th ed.). New York: Routledge.

UNFAO. (2015). The State of Food Insecurity in the World. Retrieved from http://www.fao.org/hunger/en/

7 CHAPTER 2: NARRATIVES OF CONFLICT IN AGRICULTURAL BIOTECHNOLOGY POLICY IN INDIA: THEORY, CASE STUDY, RESEARCH GOALS, AND METHODS

The Narrative Policy Framework (NPF) is a relatively new policy framework (Jones &

McBeth, 2010) compared to other policy theories and frameworks (such as Advocacy Coalition

Framework, Multiple Streams Theory, Punctuated Equilibrium Theory) but has been widely tested over the past decade in diverse policy contexts to further our understanding of the policy process (Shanahan, Jones, McBeth, et al., 2018). Its concepts are being more precisely defined

(Merry, 2016; Schlaufer, 2016; Smith-Walter, Peterson, Jones, & Nicole Reynolds Marshall,

2016), new data sources are being used (Merry, 2016), and innovative methodologies (Gray &

Jones, 2016; Weible, Olofsson, Costie, Katz, & Heikkila, 2016) are being applied to study the

NPF. NPF applications outside the United States reveal “the transportability of the NPF to diverse political systems and contextually nuanced policy domains” (Shanahan et al., 2018, p.

175). Thus, the groundwork for expanding the scope of NPF by testing its transportability and generalizability has been laid in a diversity of policy contexts.

This dissertation contributes to the endeavor of pushing the NPF forward by refining the definition of a policy narrative and examining the role of incomplete narratives (Chapter 3) and testing the transportability of narrative elements outside of the English language through its application to a linguistic context not studied before (Chapter 4). It also sheds light on the critical role of information in agricultural biotechnology policy such as evidence, risks, and benefits through an exploration of setting and plot in policy narratives (Chapter 5). This chapter briefly situates the NPF within the field of public policy scholarship outlining its historical development and main features. The case study is then discussed, followed by research goals and methods.

8 Theoretical Orientation This section provides a brief introduction to the theoretical foundation relevant to this dissertation. In the chapters that follow, the areas of literature most relevant to the specific research goals of each chapter will be presented in greater detail.

The Narrative Policy Framework (NPF)

One of the primary aims of theories of public policy is to explain the process of policy change. Given this, they attempt to address questions such as the following – how public understanding of policy problems changes, how the policy process is characterized by a period of stasis followed by sudden punctuations, and how coalitions and actors influence the policy process. Factors that influence the policy process such as resources, coalitions, punctuations, and issue definition have been studied by the leading theories of policy change such as Punctuated

Equilibrium Theory, Advocacy Coalition Framework, and Multiple Streams Approach

(Baumgartner & Jones, 1993; Kingdon, 2003; Zahariadis, 2014).

NPF contributes to theories of the policy process by focusing on the power of narratives, which had until then been largely relegated to the postpositivist realm (Jones & McBeth, 2010;

McBeth, Jones, & Shanahan, 2014a; Shanahan, Jones, McBeth, & Lane, 2013; Shanahan, Jones,

McBeth, et al., 2018). It starts with the assertion that “the power of policy narratives is something worth understanding” because (1) “policy debates are necessarily fought in the terrain of narratives, constituted by both formal institutional venues (e.g., floor debates and testimonies…) and informal venues (e.g. media, interest group websites, Twitter, YouTube, blogs) and because (2) “narratives are often asserted to affect the policy process at different times—policy decisions implementation, regulation, evaluation, and so forth”. Thus, its main contention is that “understanding the role of narratives is critical to understanding the policy process” (Shanahan et al., 2018, p. 173; italics in original).

9 McBeth and Shanahan (2004) made one of the first attempts at integrating policy narratives within policy process literature wherein they investigated policy narratives as a form of policy marketing and connected changes in policymaking to trends such as consumerism.

McBeth, Shanahan, Arnell, and Hathaway (2007) integrated narrative policy analysis into theories included under the policy change literature such as ACF. They systematically studied policy narratives to explain policy change and the role of different groups in bringing about policy change or in maintaining the status quo. Building on these earlier attempts to integrate narrative policy scholarship into policy change literature, Jones and McBeth (2010) introduced the Narrative Policy Framework (NPF) as a systematic empirical approach to the study of policy narratives that is “quantitative, structuralist, and positivist” (p. 330).

Over the past decade, there has been a proliferation of NPF focused studies. NPF focuses on three levels of analysis: micro-level studies focus on the individual as a unit of analysis, meso-level studies focus on the policy subsystem as a unit of analysis, and macro-level studies focus on the institutional or societal scale. There are four narrative core elements studied in the

NPF: setting (policy problems situated in a specific policy context), characters (heroes, villains, victims, beneficiaries, among others), plot (arc of action), and moral of the story (policy solution)

(Shanahan, Jones, McBeth, et al., 2018)1.

A policy narrative is usually built around some stated policy problem (Shanahan et al.,

1 This citation reflects the latest NPF theory but is built on previous NPF works. See Jones & McBeth, 2010, McBeth, Jones, & Shanahan, 2014, Shanahan, Jones, McBeth, & Lane, 2013, Shanahan, McBeth, & Hathaway, 2011.

10 making and often calls for policy change or transformation. Stone (2011) observes that when a policy problem is defined, one needs to also look at “how that definition defines interested parties and stakes, how it allocates the roles of bully and underdog, and how a different definition would change power relations” (p. 247).

Anderson, & Rose, 2012). A solution that aims to control the policy outcome would enable a narrative to move beyond critique or argument (Jones, 2013). Policy solutions are not always present in a narrative and such narratives may instead highlight uncertainty (Jones, Shanahan, &

McBeth, 2014).

Characters in policy narratives serve to persuade audiences and have traditionally been categorized as victims (harmed by the problem), villains (causing the problem), and heroes

(provide relief from harm/presume to have a solution) (Shanahan, Jones, McBeth, et al., 2018).

But recent NPF research has explored additional character types including beneficiaries (Weible et al., 2016), allies and opponents (Merry, 2016), and entrepreneurs and charismatic experts

(Lawton & Rudd, 2014). Beneficiaries are an important character type, particularly when characters advocate for solutions that benefit a portion of the population.

11 contested or the focal point of the policy narrative” (Shanahan et al., 2018, p. 176). NPF research on the use of evidence indicates that evidence may reflect diverging policy beliefs among coalitions (McBeth, Shanahan, Hathaway, Tigert, & Sampson, 2010; Shanahan et al., 2013;

Shanahan, McBeth, Hathaway, & Arnell, 2008); may be used in support of solutions, though it is rarely provided (Shanahan et al., 2013, 2008); may be used in different narrative strategies to indicate scientific uncertainty and disagreement (Gupta, Ripberger, & Collins, 2014; McBeth et al., 2007); may be linked to all narrative elements (Schlaufer, 2016); may be categorized further

(Smith-Walter et al., 2016); and that different types of evidence may be used at specific points in the policy process to improve policy outcomes (Mosley & Gibson, 2017).

Risk has only very recently been studied within the NPF and only within the hazards and disaster policy domain. Assessing how individuals respond to science-based risk messages and narrative-based risk messages with the same science language, Raile et al. (2018) put forth a narrative-based risk communication framework wherein they examined the mechanisms involved in narrative persuasion for risk communication. Linking NPF concepts of setting, plot, and moral of the story to risk framing, Lawlor and Crow (2018) investigated narrative construction of chronic versus urgent risk and its influence on policy discussions. The most recent NPF scholarship continues to maintain that a policy narrative features a minimum of one character and a public policy referent but acknowledges that policy scholars can define narratives with different parameters (Shanahan, Jones, McBeth, et al., 2018). The present dissertation adheres to the definition of a policy narrative cited here.

Generalizability and Transportability of the NPF

A majority of NPF studies have focused on policy contexts within the United States. A significant test for the generalizability of a particular framework is its potential application to a

12 multiplicity of policy subsystems in a variety of national and linguistic contexts. This process began with some of the first attempts at systematically applying the NPF methodology outside the United States (see, for example, Gupta et al., 2014; Lawton & Rudd, 2014; Nakyam, 2014;

Park, 2014; Schlaufer, 2016; Weible et al., 2016). Such applications might well reveal some challenges that NPF research within the United States has not yet encountered. Further applications of NPF in different policy subsystems and national and linguistic contexts would help deal with possible challenges that arise.

Given this, the present study applies the Narrative Policy Framework to the agricultural biotechnology policy process in India using a mixed-method approach (described further below) that uses data from English and Hindi media and interviews with stakeholders in addition to policy process documents for background and case study information.

Research Design: Case Study and Methodology This dissertation uses mixed methods drawing on quantitative and qualitative content analysis and qualitative interview data to empirically analyze how stakeholders interact and advocate for policy change. Using the theoretical framework discussed in the previous section, this dissertation investigates the controversial issue of agricultural biotechnology policy in India.

In particular, it focuses on the commercialization process of a genetically modified variety of eggplant called Bt eggplant. This section provides a discussion of the case study examined for this project. As discussed in the introduction, since these chapters were written as independent manuscripts focused on developing a particular aspect of the Narrative Policy Framework, each individual chapter outlines relevant case study information used for that portion of the study.

Hence, some of the information provided below is repeated in the following chapters as needed.

Agricultural Biotechnology Policy

13 Agricultural biotechnology is a broad term defined by the US Department of Agriculture

(USDA) as “a range of tools, including traditional breeding techniques, that alter living organisms, or parts of organisms, to make or modify products; improve plants or animals; or develop microorganisms for specific agricultural uses. Modern biotechnology today includes the tools of genetic engineering” (USDA, n.d.). This study focuses on the narrow aspect of genetic engineering but uses the broad terms of ‘agricultural biotechnology’ and ‘genetically modified’ crops. The term ‘genetically modified’ or GM crops is more commonly used among stakeholders and in the media in India based on media searches. Agricultural biotechnology has been a matter of contention globally since it was first adopted in the United States in 1996. Two decades later, the issue continues to cause conflict among scientists, farmers, industry, consumers, and activists, not only in the United States but also globally.

Though controversial, GM crops remain the fastest adopted crop technology and have expanded beyond the well-known four GM crops—corn, soybean, cotton, canola—to include alfalfa, sugar beets, papaya, squash, eggplant, potatoes, and apples. As per a 2017 report, the

United States continues to lead in the adoption of GM crops with ten crops planted in 2017, while Canada planted six. In the European Union (EU), only Spain and Portugal plant a transgenic variety of maize, the only GM crop approved in the EU (ISAAA, 2017). Czech

Republic and Slovakia discontinued GM maize due to higher demand for non-biotech crops

(ISAAA, 2017). Thus, EU continues to lead the Western Hemisphere in rejection of agricultural biotechnology. Commercialization of GM crops in Africa has increased with twelve GM crops in thirteen countries. Ten countries in Latin America planted GM crops in 2017 contributing to

42% of the global biotech area. Leading biotech countries in Asia and the Pacific region include

India, followed by Pakistan, China, Australia, the Philippines, Myanmar, Vietnam, and

14 Bangladesh contributing to 10% of the global biotech crops (ISAAA, 2017). Lack of government approval has blocked several GM crops such as Golden Rice, Bt eggplant, and drought tolerant and insect resistant maize in different regions. Bt eggplant, commercialized in Bangladesh is banned in neighboring India. Golden Rice faces stiff opposition in developing countries in Asia,

Latin America, and Africa which have high incidence of vitamin A deficiency. Regulatory approval for drought tolerant and insect resistant maize, a public-private collaboration of four countries in sub-Saharan Africa remains stalled, while Brazil approved a virus-resistant bean in

2011 but the technology has not reached farmers (ISAAA, 2017).

There is no robust explanation for the variation in reception of GM crops—some countries like the United States and Canada accept agricultural biotechnology with little contention (though state policies show variation here as well), while others like those in the

European Union have switched positions over time and other countries like Russia have rejected the technology altogether (Herring & Paarlberg, 2016; Paarlberg, 2001).

Agricultural Biotechnology Policy in India: The Case of Bt Eggplant

India is an excellent country in which to study the GM discourse. After the green revolution in the 1960s, India became largely self-sufficient in agricultural production (Das &

Bhardwaj, 2015). However, rapid growth in population, reduction in arable land due to industrialization, and reliance on monsoons are some of the reasons leading to reduced agricultural production. Agricultural biotechnology has increasingly been publicized as the best means of bringing about a second green revolution (Restall, 2014). However, the matter remains contentious as will be detailed below.

15 On 26th March 2002, the regulatory body for agricultural biotechnology in India, Genetic

Engineering Approval Committee2 (GEAC), recommended the commercialization of a transgenic variety of cotton called Bt cotton ushering the GM debate into the Indian public sphere (Kalle & Ejnavarzala, 2016). India is one of the top five countries with area under GM crop cultivation with Bt cotton as the only transgenic crop approved for commercial cultivation

(ISAAA, 2017). Despite the widespread cultivation of Bt cotton in India, other crops have not received approval for commercial cultivation.

After the widespread adoption of Bt cotton, Bt eggplant was being perceived as potentially the first food crop to be considered seriously by the regulatory system (Ramaswami

& Pray, 2007). In India, there was intense public debate nationwide regarding its commercialization. This transgenic that is created through the insertion of a gene cry1Ac from the soil bacterium (Bacillus thuringiensis) into eggplant is said to provide the plant with resistance against lepidopteran insects like the Fruit and Shoot Borer (Leucinodes orbonalis) and

Fruit Borer (Helicoverpa armigera). Eggplant is a popular vegetable crop in India among small scale farmers as well as among low income consumers making it known as the “poor man’s vegetable crop.” It is low in calories, has a high nutritional value and water content, serves as a good source of fiber, calcium, phosphorus, folate, vitamins B and C, and is used in traditional medicine, whereas the plant itself is used as fuel (Bandopadhyay, Sinha, & Chaudhary, 2012, p

238). After nine years of testing by a complex array of state institutions that were coordinated by the GEAC including seven government departments, committees and institutes, Bt eggplant’s hybrids and varieties were approved in October 2009 (Bagla & Stone, 2013; Herring, 2012,

2015). However, in 2010, it could no longer be planted legally after Minister for Environment

2 It was renamed as the Genetic Engineering Appraisal Committee in 2010.

16 and Forests at the time rejected GEAC’s decision and placed an indefinite moratorium pending further evaluations (Bandopadhyay et al., 2012; Herring, 2012, 2015). The moratorium overturned the regulatory body’s decision to allow the commercialization of Bt eggplant.

The moratorium on Bt eggplant is often cited by GMO-opponents as a victory in support of their continued opposition to any transgenic crop including GM mustard which has been in the policy pipeline in India for several years now. The controversy surrounding India’s second transgenic crop and first transgenic food crop was intense and focused on risk. Even though Bt eggplant carried the same transgene for insect resistance as Bt cotton (cry1Ac) and underwent the same regulatory process, its risk assessment for regulation was controversial (Gupta, 2011).

With Bt cotton, state and cultivator interests dominated (see Herring, 2015 for detailed discussion on Bt cotton controversy) but with Bt eggplant the politics of risk dominated, especially given the susceptibility of food crops to “anxiety framings” (Herring & Paarlberg,

2016, p 410).

The Bt eggplant policy process in India provided a unique opportunity for participatory decision-making. Following a public outcry in the wake of the GEAC approval of Bt eggplant in

2009, the Minister for Environment held a series of stakeholder consultations nationwide that were attended by farmers, civil society organizations, consumers, scientific experts, industry, environmental groups, government officials among others. The Minister gathered evidence from multiple sources to enlarge the evidence base including views of state and regional governments and independent scientists from India and abroad. On the basis of his review of evidence from multiple sources, he imposed a moratorium on the commercialization of Bt eggplant “till such time independent scientific studies establish, to the satisfaction of both the public and professionals, the safety of the product from the point of view of its long-term impact on human

17 health and environment” (Ministerial Note, MoEF 2010:17).” (qtd. in Kalle & Ejnavarzala, 2016, p 32; Herring, 2015; Ramesh, 2015).

Research Goals and Findings Using the theoretical lens of the NPF, agricultural biotechnology policy in India provides a rich case study to examine policy narratives in a distinct policy subsystem. Each of the empirical chapters expands the scope of the NPF and moves it forward by: refining the definition of a policy narrative and examining the role of incomplete narratives (Chapter 3); by testing the transportability of narrative elements outside of the English language through its application to a linguistic context not studied before (Chapter 4); and by refining our understanding of setting and plot in policy narratives through an examination of how evidence of risks and benefits is used by stakeholders in the policy process (Chapter 5). Below are the abstracts for the three

“data-chapters” (Chapters 3-5), which highlight the research goals and findings of each chapter:

Chapter 3 – An Examination of Policy Narratives in Agricultural Biotechnology Policy in

India

The Narrative Policy Framework (NPF) focuses attention on the importance of narratives in policy debates and on their empirical analysis. While NPF has become an increasingly important and accepted approach to studying the policy process, the vast majority of research applies it to the policy contexts of the United States, which limits tests of its potential generalizability and responsiveness to cultural specificity. To broaden the contextual scope of the approach, this study applies the NPF to a non-U.S. policy context through examining the controversial issue of agricultural biotechnology policy in India. It analyzes media coverage from leading English newspapers in India to explore the strategic use of narrative variables in policy narratives. In doing so, it highlights the important role of incomplete policy narratives in policy debates and outcomes. Policy narratives do not always contain a full suite of narrative components, and yet

18 they may be among the most common messages received by the public and political actors.

Through an analysis of incomplete narratives, this study attempts to further refine the definition of policy narratives and consider which narratives are important from empirical and audience reception perspectives. Results show that incomplete narratives occur more frequently and contain relevant narrative variables.

Chapter 4 - Policy Narratives across Two Languages: A Comparative Study using the

Narrative Policy Framework

The Narrative Policy Framework (NPF) focuses on the role of narratives, drawing from a rich scholarship in narrative, language, and culture. Despite the understanding that narratives are constructions of language and that narrative construction differs across languages, NPF studies have not focused on policy narratives in languages other than English. Language characteristics of narratives are important to assess the stability of policy narratives when there are multiple dominant languages in a political system. This study investigates the use of narrative elements in policy narratives in agricultural biotechnology policy in India across Hindi and English media coverage to examine the NPF assumption that narratives have generalizable narrative elements irrespective of variation in linguistic context and, specifically, tests the transportability of narrative elements. Findings validate the transportability of narrative elements in Hindi narratives, indicating variation in the use of narrative elements over time, and have implications for applying NPF across languages.

Chapter 5 - Sources of Evidence for Risks and Benefits in Agricultural Biotechnology

Policy in India: Examining Setting and Plot in Policy Narratives

Though introduced more than two decades ago, agricultural biotechnology remains contentious.

New technologies bring uncertainty about their risks and promises and create new policy process

19 challenges for governments assessing risks and benefits in the face of contradictory evidence.

Stakeholders typically claim to have evidence to support their positions, which often focuses on risks and benefits of agricultural biotechnology. This study examines the commercialization process for a genetically modified variety of eggplant in India called Bt eggplant with a focus on the evidence in support of claims about risks and benefits in policy narratives and explores moral notions of risk as expressed in actor beliefs. Using the Narrative Policy Framework to structure the analysis and drawing on data from elite interviews and leading English and Hindi newspapers in India, it operationalizes setting and plot in policy narratives. Findings indicate that stakeholders use different sources of evidence and proponents de-emphasize risks and exclusively highlight benefits while opponents invoke multi-dimensional risk. Lastly, risk perceptions of stakeholders are influenced by moral notions of risk.

Methods: Data Collection and Analysis As recommended for case study research, multiple sources of data were collected to triangulate findings when possible, and to be transparent and clear (Miles & Huberman, 1994;

Yin, 1994). This allowed the researcher to focus on different aspects and to understand the case from various perspectives. This section briefly describes each data source and the relevant analysis methods used for this dissertation. Each empirical chapter provides further description of data sources and data analysis methods as relevant to that chapter.

Documents: First, all policy process documents were gathered for contextual background information on the policy beginning from mid-2000 when Bt eggplant was being perceived as potentially the first GM food crop to be considered seriously by the regulatory system

(Ramaswami & Pray, 2007). This provided background information relevant to the case study.

Second, media data were collected from the leading Hindi (Dainik Jagran) and English (Times of

20 India and Hindustan Times) newspapers in India and were analyzed using an NPF focused codebook using NPF coding procedures. To capture breadth of coverage, sampling timeframe was three years before and after the ban was imposed on Bt eggplant: February 9, 2007 -

February 9, 2013. This allows us to examine policy narratives while the policy was being actively discussed in the media.

For the English media dataset: A total of 1,212 articles (397 from Times of India and 815 from Hindustan Times) were downloaded using search terms: “Bt brinjal” OR “Bt eggplant” OR

“genetically modified” OR “agricultural biotechnology.” Duplicate articles or those not focusing on Bt eggplant/brinjal or not written in narrative form were removed from the dataset leading to a final English media dataset of 227 articles (87 from Times of India and 140 from Hindustan

Times) that were coded initially. Based on the ‘policy narrative’ definition (Shanahan, Jones,

McBeth, & Radaelli, 2018), media articles without a policy referent and a character were removed leading to 171 articles (Times of India - 51 and Hindustan Times - 120). For the Hindi media dataset, archives of the New Delhi edition of Dainik Jagran were accessed for the same timeframe as the English media coverage. After a manual search of the dataset, articles focusing on Bt eggplant (बीट$ बगन& ) were collected (n=20) and only policy narratives were included in the final dataset (n=19). Further description of the collection and analysis of the English media data and Hindi media is provided in Chapters 3 and 4 respectively.

Interviews: Third, semi-structured interviews were conducted in June-August 2017 as per guidelines provided by Rubin and Rubin (2005). 41 stakeholders were contacted and 23 were interviewed. Only those interview transcripts that adhered to the policy narrative definition (a character and policy referent) are included in this analysis (n=20). Interviews were conducted

21 with stakeholders in agricultural biotechnology who are currently or were formerly affiliated with non-profit organizations (n=8), industry (n=4), research organizations and public universities (n=5), and government including regulatory bodies (n=3). Since stakeholders were identified through an analysis of media coverage, their public position on the issue was known.

The interview attempted to elicit their policy narratives with a focus on specific narrative elements. To maximize consistency of coding and analysis, interview transcripts were coded using a constant comparative approach in NVivo software (Miles & Huberman, 1994). Detailed description of the collection and analysis of the interview data is provided in Chapter 5.

22 References

Bagla, P., & Stone, R. (2013). Scientists clash swords over future of GM food crops in India. Science, 340(May), 539–540. https://doi.org/10.1126/science.340.6132.539

Bandopadhyay, R., Sinha, P., & Chaudhary, B. (2012). Is Bt-brinjal ready for future food?--A critical study. Indian Journal of Biotechnology, 11(2), 238–240.

Baumgartner, F. R., & Jones, B. D. (1993). Agendas and instability in American politics. University of Chicago Press.

Gray, G., & Jones, M. D. (2016). A qualitative narrative policy framework? Examining the policy narratives of US campaign finance regulatory reform. Public Policy and Administration, 31(3), 193–220. https://doi.org/10.1177/0952076715623356

Gupta, A. (2011). An evolving science-society contract in India: The search for legitimacy in anticipatory risk governance. Food Policy, 36(6), 736–741.

Gupta, K., Ripberger, J. T., & Collins, S. (2014). The Strategic Use of Policy Narratives: Jaitapur and the Politics of Siting a Nuclear Power Plant in India. In M. D. Jones, E. A. Shanahan, & M. K. McBeth (Eds.), The Science of Stories: Applications of the Narrative Policy Framework in Public Policy Analysis (1st ed., pp. 89–106). New York: Palgrave Macmillan.

Herring, R. (2015). State science, risk and agricultural biotechnology: Bt cotton to Bt Brinjal in India. Journal of Peasant Studies, 42(1), 159–186.

Herring, R. J. (2012). State science and its discontents: why India’s second transgenic crop did not follow the path of Bt cotton’. In Weihenstephaner Socio-Economic Seminar, Center of Life and Food Sciences Weihenstephan, Technische Universität M ünchen (Vol. 13).

Herring, R., & Paarlberg, R. (2016). The Political Economy of Biotechnology. Annual Review of Resource Economics, 8, 397–416. https://doi.org/10.1146/annurev-resource-100815-095506

Jones, M. D. (2013). Cultural Characters and Climate Change: How Heroes Shape Our Perception of Climate Science. Social Science Quarterly, 95(1), 1–39. https://doi.org/10.1111/ssqu.12043

23 Jones, M. D., & McBeth, M. K. (2010). A Narrative Policy Framework: Clear Enough to Be Wrong? Policy Studies Journal, 38(2), 329–353. Retrieved from http://search.proquest.com/docview/210543073?accountid=452

Jones, M. D., Shanahan, E. A., & McBeth, M. K. (2014). The Science of Stories: Applications of the Narrative Policy Framework in Public Policy Analysis. Palgrave Macmillan.

Kalle, J., & Ejnavarzala, H. (2016). Moratorium on Genetically Modified Brinjal in India: Is Evidence-Based Policy making An Adequate Framework? Asian Biotechnology & Development Review, 18(2).

Kingdon, J. W. (2003). Agendas, Alternatives, and Public Policies (2nd ed.). Harper Collins Publishers.

Lawlor, A., & Crow, D. (2018). Risk‐Based Policy Narratives. Policy Studies Journal, 46(4), 843–867.

Lawton, R., & Rudd, M. (2014). A Narrative Policy Approach to Environmental Conservation. AMBIO, 1–9. https://doi.org/10.1007/s13280-014-0497-8

McBeth, M. K., Jones, M. D., & Shanahan, E. A. (2014a). The Narrative Policy Framework. In P. A. Sabatier & C. M. Weible (Eds.), Theories of the Policy Process (3rd ed.). Boulder, CO: Westview Press.

McBeth, M. K., & Shanahan, E. A. (2004). Public opinion for sale: The role of policy marketers in Greater Yellowstone policy conflict. Policy Sciences, 37(3–4), 319–338.

McBeth, M. K., Shanahan, E. A., Arnell, R. J., & Hathaway, P. L. (2007). The Intersection of Narrative Policy Analysis and Policy Change Theory. Policy Studies Journal, 35(1), 87– 108. Retrieved from http://search.proquest.com/docview/210546782?accountid=452

McBeth, M. K., Shanahan, E. A., Arrandale Anderson, M. C., & Rose, B. (2012). Policy Story or Gory Story? Narrative Policy Framework Analysis of Buffalo Field Campaign’s YouTube Videos. Policy & Internet, 4(3–4), 159–183. https://doi.org/10.1002/poi3.15

McBeth, M. K., Shanahan, E. A., Hathaway, P. L., Tigert, L. E., & Sampson, L. J. (2010). Buffalo tales: interest group policy stories in Greater Yellowstone. Policy Sciences, 43(4), 391–409.

Merry, M. K. (2016). Constructing Policy Narratives in 140 Characters or Less: The Case of Gun Policy Organizations. Policy Studies Journal, 44(4), 373–395. https://doi.org/10.1111/psj.12142

Miles, M. B., & Huberman, M. (1994). Qualitative data analysis: An expanded sourcebook. sage.

24 Mosley, J. E., & Gibson, K. (2017). Strategic use of evidence in state-level policymaking: matching evidence type to legislative stage. Policy Sciences, 50(4), 697–719.

Nakyam, Su. (2014). Educational Decentralization Policy in Thailand: Unpacking Its Labyrinth to Pinpoint an Appropriately Further Step. In International Conference on Public Administration.

Park, Y.-S. (2014). A Study of the construction permit process of 2nd Lotte World (skyscraper) using the Narrative Policy Framework. The Korean Governance Review, 21(2), 101–125.

Paarlberg, R. L. (2001). The politics of precaution: Genetically modified crops in developing countries. Intl Food Policy Res Inst.

Raile, E. D., King, H., Shanahan, E. A., McEvoy, J., Izurieta, C., Bergmann, N., … Poole, G. C. (2018). Narrative-based Risk Communication: A Lingua Franca for Natural Hazard Messages? In Midwest Political Science Association.

Ramaswami, B., & Pray, C. E. (2007). India: confronting the challenge–the potential of genetically modified crops for the poor. The Gene Revolution: GM Crops and Unequal Development, 156–174.

Ramesh, J. (2015). Green signals: ecology, growth, and democracy in India. Oxford University Press.

Restall, H. (2014, November 21). Growing a Second Green Revolution. The Wall Street Journal. Retrieved from http://www.wsj.com/articles/growing-a-second-green-revolution- 1416613158

Schlaufer, C. (2016). The Narrative Uses of Evidence. Policy Studies Journal, 46(1), 90–118. https://doi.org/10.1111/psj.12174

Shanahan, E. A., McBeth, M. K., Hathaway, P. L., & Arnell, R. J. (2008). Conduit or contributor? The role of media in policy change theory. Policy Sciences, 41(2), 115–138.

Shanahan, E. A., Jones, M. D., McBeth, M. K., & Lane, R. R. (2013). An Angel on the Wind: How Heroic Policy Narratives Shape Policy Realities. Policy Studies Journal, 41(3), 453– 483. https://doi.org/10.1111/psj.12025

Smith-Walter, A., Peterson, H. L., Jones, M. D., & Nicole Reynolds Marshall, A. (2016). Gun Stories: How Evidence Shapes Firearm Policy in the United States. Politics & Policy, 44(6), 1053–1088. https://doi.org/10.1111/polp.12187

25 Stone, D. (2011). Policy Paradox: The Art of Political Reason. New York, NY: WW Norton and Company.

USDA. (n.d.). No Title. Retrieved August 25, 2017, from https://www.usda.gov/topics/biotechnology/ biotechnology-frequently-asked-questions-faqs

Weible, C. M., Olofsson, K. L., Costie, D. P., Katz, J. M., & Heikkila, T. (2016). Enhancing Precision and Clarity in the Study of Policy Narratives: An Analysis of Climate and Air Issues in Delhi, India. Review of Policy Research, 33(2).

Yin, R. (1994). Case study research. Design and methods. red., Thousand Oaks CA, Sage publications.

Zahariadis, N. (2014). Ambiguity and Multiple Streams. In Paul A. Sabatier & C. M. Weible (Eds.), Theories of the Policy Process (3rd ed.). Boulder, CO: Westview Press.

26 CHAPTER 3: AN EXAMINATION OF POLICY NARRATIVES IN AGRICULTURAL BIOTECHNOLOGY POLICY IN INDIA3

Public policy scholarship attempts to understand how knowledge and information are used in the policy process. This knowledge and information is often communicated through policy narratives that are specifically constructed about policy issues and tell stories that contain narrative elements and strategies (McBeth, Jones, & Shanahan, 2014b). These policy narratives are the focus of the Narrative Policy Framework (NPF) which provides theoretical and empirical tools for analyzing the role of narratives in the policy process (Jones & McBeth, 2010). In particular, NPF scholarship focuses on variations in content of narratives through a systematic study of several narrative elements as well as narrative strategies and belief systems.

NPF’s central questions revolve around the empirical role of policy narratives in the public policy process, and whether policy narratives influence policy outcomes (Shanahan et al.,

2013). The present chapter focuses on the former question as it explores the role of narratives. It expands scholarship on the NPF by exploring connections between narrative variables and policy outcomes in a non-U.S. policy context. It examines the controversial issue of agricultural biotechnology policy in India. With a focus on how narrative variables are used in policy narratives, it analyzes media coverage from two leading English newspapers in India and provides an in-depth discussion of how narrative variables appear in policy narratives generated by the media. Further, it examines the strategic use of narrative variables in the media across losing, winning, and incomplete narratives with specific attention to certain NPF variables.

Policy narratives do not always contain a full suite of narrative components and yet these

3 This chapter is a version of the following article and should be cited as follows: Huda, J. (2018). An Examination of Policy Narratives in Agricultural Biotechnology Policy in India. World Affairs, 181(1), 42–68.

27 incomplete narratives may be among the most common messages received by the public and political actors. Through an analysis of incomplete narratives, this study attempts to further refine the definition of policy narratives and consider which narratives are important, both from empirical and audience reception perspectives.

The study of policy narratives using NPF has expanded beyond the United States to countries such as Switzerland, United Kingdom, India, Thailand, and Korea (Gupta et al., 2014;

Lawton & Rudd, 2014; Nakyam, 2014; Park, 2014; Schlaufer, 2016; Weible et al., 2016). NPF studies in the Indian context have investigated strategic use of policy narratives by coalitions surrounding the Jaitapur Nuclear Power Plant and developed a network-based approach for coding characters in the analysis of air and climate issues in Delhi, respectively (Gupta et al.,

2014; Weible et al., 2016). This study attempts to contribute to, and expand further, this growing body of comparative NPF scholarship as an important next step in expanding our understanding of policy narratives, which intuitively are embedded with significant amounts of culturally specific content. Because of this, studies beyond U.S. policy processes are vital to conduct.

This article briefly presents literature on NPF and its contribution to public policy scholarship. It then provides a discussion on how media content has been studied under NPF and its role in the policy process. I subsequently make a case for the inclusion of incomplete policy narratives before defining winning and losing narratives. Research methods are then discussed followed by this study’s findings and discussion. I conclude by extrapolating some of the key implications of the study, directions for future research, and a note on the limitations facing this and similar studies.

NPF and Public Policy Scholarship

28 NPF considers narratives as central to the policy process and as extremely important to shape the conduct and outcomes of all aspects of government since they may potentially provide information on dynamics, beliefs, and actor behavior within the policy process (Jones, 2013;

Jones & Jenkins‐Smith, 2009; Jones & McBeth, 2010; Shanahan, Jones, & McBeth, 2011;

Shanahan et al., 2013; Shanahan, McBeth, & Hathaway, 2011; Weible & Schlager, 2014).

Narratives are critical to the meaning-making process since they can be used to persuade actors toward a particular policy preference or to influence decision making (Jones, Shanahan, et al.,

2014). Analysis of policy narratives enables scholars to gain insight into the dynamics of public policy issues, the opinions of stakeholders who advocate within policy subsystems, and possible directions with regard to policy decisions (Weible et al., 2016).

NPF focuses on three levels of analysis: micro-level studies focus on the individual as a unit of analysis; meso-level studies focus on the policy subsystem as a unit of analysis; and macro-level studies focus on the institutional or societal scale (Shanahan, Jones, et al., 2011).

Meso-level NPF studies have often relied on public consumption documents distributed by actors from advocacy organizations. These documents are generated and disseminated with policy- relevant intent to persuade actors to enact change. However, narratives in media may not always be disseminated with policy-relevant intent and may appear to be incomplete to the NPF researcher. However, they are an important source of information for the public and hence need to be studied as policy narratives in the policy process (Crow et al., 2016). A more detailed explanation on the role of media in the policy process and the need to consider incomplete policy narratives is provided in the following section. In addition to the narrative core elements (setting, characters, plot, and solution), this study includes two more components: use of scientific evidence and discussion of risks and benefits.

29 Previous NPF studies have examined the use of evidence in narratives as an indication of diverging policy beliefs among coalitions (McBeth et al., 2010; Shanahan et al., 2013, 2008).

Use of evidence has also been studied as a narrative strategy in the context of scientific uncertainty and disagreement (Gupta et al., 2014; McBeth et al., 2007) while Schlaufer (2016) has linked scientific evidence to all narrative elements in an attempt to systematize how evidence is used in narratives. Thus, it is important to study the use of scientific evidence in a narrative.

On the other hand, risk perceptions have not been as widely studied in the NPF. These may be essential to understanding how policy problems are defined particularly in scientific issues that are fraught with risk. In an NPF study focused on wildfires in Colorado, Crow and others (2016) studied risk perceptions in narratives of natural hazards as these may influence policy decisions that aim to reduce risk from future or current hazards. Both of these components are relevant to include in a discussion on agricultural biotechnology policy since the science behind agricultural biotechnology and its risks and benefits have been intensely debated and these components help provide a thematic framing that connects a single incident to larger societal trends, problems, or causes (Crow et al., 2016; Iyengar, 1990).

NPF and Policy Narratives in the Media

Actors involved in the policy process use strategically constructed stories to communicate with, persuade, or influence the public. “[These stories] come in many forms and from many stakeholder sources—elected officials give speeches, interest groups write newsletters and press releases, concerned citizens write letters to the editor, and the media writes editorials and news stories” (Shanahan, McBeth, et al., 2011, p 374). All of these stories contain varied narrative components that are connected to actor beliefs and policy preferences. However, these stories become policy narratives only “when the author or group strategically constructs the

30 story to try to win the desired policy outcome” (p 375). This has also been referred to as a

“policy stance” present in the policy narrative. Policy narratives come in many forms and from different sources. Those that originate from interest groups or elected officials are more likely than those from media to include a policy stance as these stakeholders have vested interests in promoting their policy preference. Equating narratives constructed by stakeholders and by media can lead to a flawed analysis, particularly because media narratives may not always promote a policy preference but may still contain requisite narrative components.

Examining definitions of policy narratives as they have appeared in the NPF literature over time may help to clarify this issue further. One of the earliest definitions of a policy narrative appears in Jones & McBeth (2010, p 340) which outlines its minimum qualities: (1)

Setting or a context, (2) Plot, (3) Characters, and (4) Moral of the story. A policy stance or preference on the issue may have been assumed but is not mentioned as a requirement. More recent NPF treatments state that a policy narrative must have “at least one character and some reference to a public policy preference or stance” (McBeth et al., 2014, p 229), but also assert that, “a policy narrative will have a minimum of one character and a referent to the public policy of interest (e.g., problem, solution, evidence for, etc.)” (Jones et al., 2014, p 7). The examples in the parentheses do not refer to a policy stance as suggested by the first definition. It further states that, “it is possible that a communication would be considered a policy narrative without a solution” (Jones et al., 2014, p 7). A policy stance usually manifests as a solution, but that is not included in this definition. Thus, the requirement of the inclusion of a policy stance varies based on these definitions.

Recent NPF research delves into the role of media in the policy process and acknowledges that “Media, of course, are not the same, and as such they have different

31 influences on the audiences they reach” (Crow & Lawlor, 2016, p 479). Arnold (1990) distinguished between two policy spheres where the work of policy making occurs: the visible and the invisible realm. Media contribute to the formation of public opinion in the visible realm and this may indirectly put constraints on policy makers depending upon public opinion and media coverage. On the other hand, in the invisible realm, the media do not pay attention and neither does the public, which remains ignorant. Policy makers are able to make choices unconstrained by public opinion or public scrutiny within the invisible realm.

In the visible realm, media’s interaction with the policy process occurs in two important ways: through the selection of issues that are to be highlighted for the public and policy makers

(i.e., the topical focus of narratives); and through the way meaning is attached to the issue to make it understandable (i.e., the strategic construction of narratives). Media may not always strategically construct the narratives and may simply highlight issues for the public and policy makers. Thus it may provide information to the public in an unbiased manner, but also play an active role in changing opinions and influencing agendas through a strategic framing and construction of policy narratives (Crow & Lawlor, 2016; McLeod, Kosicki, & McLeod, 2002).

Media play a role in the policy process “by constructing (or co-constructing with policy advocates) the images used to communicate about and understand policy issues… framing issues in certain ways… and disseminating the narratives communities use to discuss problems, policies, and solutions…” (Crow et al., 2016, pp 5-6).

NPF research demonstrates that media play a role in the production of policy narratives but it remains unclear “whether media actors are considered active, voluntary members of coalitions, or whether they are incidental members of such coalitions through their professional roles” (Crow & Lawlor, 2016, p 474). In case of print media in particular, “the audience is less

32 likely to have a predetermined positive or negative stance toward reading a news article as opposed to an interest group’s or political official’s narrative (Entman, 1995). Whereas readers are prepared for political officials or interest groups to put forth policy outcomes, the mainstream media is presumed by general audiences to be more neutral in their reporting” (Shanahan,

McBeth, et al., 2011, p 376). Thus, not all media narratives may constitute policy narratives given the neutrality of the narrative. Media accounts can often outline policy issues with none or only some of the narrative elements present in them. Yet, “readers are most likely to be open to ingesting [policy narratives in media accounts] for information” (p 393). Shanahan and others

(2008) tested whether the media serve as conduits or contributors. Their results demonstrate “a more complex policy landscape, perhaps necessitating a move away from a dichotomous conceptualization” about contributor or conduit to “matters of degree” wherein media can be contributors in some instances and conduits in others (p 130).

In addition to being conduits or contributors, media accounts may contain information on policy issues from multiple perspectives. Although they may not be ‘policy narratives’ in the sense that they do not advocate for a specific policy preference, they may yet contain the required narrative elements (policy referent and character) (Jones et al., 2014, p 7). Policy narratives may at times try to provide holistic information on the policy issue rather than shape public opinion toward a specific policy goal. Even though these narratives may contain the policy referent and character components of a narrative, these may tend to be excluded from an

NPF analysis given that they do not outline a specific stance on the issue. Since media can act in multiple roles of conduits and contributors—as well as disseminators of multiple sides of the issue—this study undertakes a classification of narratives as not only losing and winning policy narratives akin to much NPF scholarship, but also ‘incomplete’ policy narratives. The

33 ‘incomplete’ policy narrative is one which does not advocate for a specific policy preference and contains multiple views from both sides with regard to the policy issue. All of these narratives contain the required two narrative elements of policy referent and character.

Lastly, analyzing a policy stance or judgement often occurs when the study is directly examining winning or losing coalitions. For example, Weible and others (2016) conducted an

NPF analysis on air and climate issues in Delhi, India but did not focus on ‘losing’ and ‘winning’ coalitions per se, thereby not needing to code for a policy stance. The suggestion here is that the presence of a policy stance—though arguably important in NPF research—may not always be central and included in a policy narrative, depending upon the focus of the NPF study. Thus, not every NPF study may code for a policy stance. Following this lead, the present research therefore includes ‘incomplete’ policy narratives: i.e., policy narratives that do not advocate for a particular policy preference but contain the requisite narrative elements of referent and character to qualify for a policy narrative.

NPF research has primarily analyzed winning and losing narratives in the context of narrative strategies used by winning and losing coalitions (Schattschneider, 1960) and the corollary perceptions of those coalitions regarding their status as winners or losers. Sometimes, clear winners and losers emerge in a policy debate (Crow & Berggren, 2014; Shanahan et al.,

2013) whereas in intractable policy issues they may not emerge (McBeth et al., 2007). Therefore, definitions of winning and losing narratives need to align with this aspect. For the present study, a clear winner and loser emerged, enabling the classification of narratives not in favor of the policy as winning narratives and those in favor as losing narratives.

Broadening the NPF: Applications in Non-U.S. and Other Policy Contexts

34 A majority of NPF studies have focused on policy contexts within the United States, but there is a dearth of NPF studies in contexts outside that country. A significant test for the generalizability of a particular framework is its potential application to a multiplicity of policy subsystems in a variety of national contexts. This process has recently begun with some of the first attempts at systematically applying the NPF methodology outside the United States (see e.g., Gupta, Ripberger, & Collins, 2014; Lawton & Rudd, 2014; Nakyam, 2014; Park, 2014;

Schlaufer, 2016; Weible, Olofsson, Costie, Katz, & Heikkila, 2016). Such applications might well reveal some challenges that NPF research within the United States has not yet encountered.

Meso-level NPF research has relied on content analysis of public consumption documents that have been relatively accessible in the United States. However, both in Gupta et al. (2014) and in

Weible et al. (2016) studies (n = 55 and 75 respectively), they were able to access relatively few public consumption documents, likely because coalitions in countries like India tend to distribute fewer such documents. Also, there are discrepancies in terms of how coalitions use the Internet to distribute these documents, which complicates their accessibility. In contexts where large-n studies cannot be conducted, other approaches may need to be developed for meso-level analysis. This is one area that remains underdeveloped in narrative scholarship. Further applications of NPF in non-U.S. contexts would help to deal with possible challenges that arise.

Given this, the present study involving a case study of adoption of genetically modified crops in

India makes a useful step in expanding the scope of NPF.

Agricultural Biotechnology Policy—The Case of Bt Eggplant in India

Agricultural biotechnology is a broad term defined by the United States Department of

Agriculture (USDA) as “a range of tools, including traditional breeding techniques, that alter living organisms, or parts of organisms, to make or modify products; improve plants or animals;

35 or develop microorganisms for specific agricultural uses. Modern biotechnology today includes the tools of genetic engineering” (USDA, n.d.). This study focuses on the narrow aspect of genetic engineering but uses the broad terms of ‘agricultural biotechnology’ and ‘genetically modified’ (GM) crops. The term ‘genetically modified’ crops is more commonly used among stakeholders and in the media in India based on media searches. Agricultural biotechnology has been a matter of contention globally since it was first adopted in the United States in 1996. Two decades later, the issue continues to cause conflict among scientists, farmers, industry, consumers, and activists, not only in the United States, but globally. To date, there have been no definite findings regarding the variation in reception of GM crops among nations and stakeholders (Herring, 2006, 2015). This study uses NPF to understand the varied narratives and examines their role in the policy process.

India is a crucial case study, given its trajectory—a long history of promoting agricultural biotechnology since the mid-1980s, the adoption of a GM crop (Bt cotton) in 2002, and a ban in

2010 on the first GM food crop (Bt eggplant or Bt brinjal, as it is commonly known in India). In

February 2016, the Indian government was expected to make a decision on allowing GM mustard. However, that decision has been deferred (Damodaran & Sinha, 2016; Mohan, 2016).

In mid-2000, Bt eggplant was potentially the first GM food crop to be considered seriously by the regulatory system (Ramaswami & Pray, 2007). There was intense public debate nationwide regarding its commercialization. It is the first locally developed GM food crop in India created by Maharashtra Hybrid Seeds Company (Mahyco), a joint venture with Monsanto, the St. Louis– based seed giant (Centre For Environment and Education, 2010; Jayaraman, 2010). After nine years of rigorous testing by a complex array of state science institutions that were coordinated by the Genetic Engineering Approval Committee (GEAC)—including seven government

36 departments, committees, and institutes—Bt eggplant’s hybrids and varieties were approved in

October 2009 (Bagla & Stone, 2013; Herring, 2012, 2015). However, in 2010, the Minister for

Environment and Forests at the time, rejected GEAC’s decision and placed an indefinite moratorium owing to strong public opposition (Bandopadhyay et al., 2012; Herring, 2012, 2015).

He stated that the moratorium would remain until studies prove “the safety of the product from the point of view of its long-term impact on human health and environment” (Bagla, 2010, p

767). Using the commercialization process of Bt eggplant as the policy issue under consideration, this study examines variation in policy narrative elements across winning, losing, and incomplete narratives.

Research Methods

Data for the study were collected from two of the leading English newspapers in India:

Times of India and Hindustan Times.4 Policy documents were first gathered to provide contextual policy information such as a timeline, actors involved, and policy issue information. To keep the analysis more focused and to measure content of policy narratives over a significant period of time, the sampling timeframe was chosen as three years before and after the original policy decision to ban Bt eggplant in India: February 9, 2007 to February 9, 2013. Online archives of the New Delhi edition of Times of India and Hindustan Times were accessed through the

ProQuest News and Newspapers database. The search terms, daily circulation, and article counts are included in Table 3.1.

4 As per the Indian Readership Survey (2014), Times of India and Hindustan Times are the top two English daily newspapers in India.

37 Table 3. 1: Search Terms, Newspapers, and Article Counts

Newspaper Audience Circulation Search Terms Article Counts Times of National 7,590,000 daily Bt brinjal, Bt eggplant, genetically 87 India modified, agricultural biotechnology Hindustan National 4,515,000 daily Bt brinjal, Bt eggplant, genetically 140 Times modified, agricultural biotechnology Total 12,105,000 227

1,212 articles (397 from Times of India and 815 from Hindustan Times) were downloaded using search terms: “Bt brinjal” OR “Bt eggplant” OR “genetically modified” OR

“agricultural biotechnology.” Articles that were duplicates, did not focus on Bt eggplant, or were not written in narrative form (such as lists or bulletins) were removed, which led to a final count of 227 articles (87 from Times of India and 140 from Hindustan Times) coded for this study.

Since this study focuses on how the use of narrative elements varied across the three types of narratives, it is important to consider their frequency across the six-year timeframe.

Figure 3.1 shows the coverage of losing, winning, and incomplete narratives over the six-year period. Incomplete narratives had the highest frequency (48.21 percent, n=81), followed by winning narratives (38.69 percent, n =65), and losing narratives were the least (13.10 percent, n=

22). Media provided more incomplete narratives at the time of the policy decision and also actors supporting a ban on Bt eggplant were more active and prevalent in the visible realm (where media contribute to formation of public opinion). Given the higher frequency of incomplete narratives, compared to winning and losing narratives, makes it more salient to include them in the analysis.

38 Figure 3. 1: Coverage of Types of Narratives

Types of Narratives 35 30 25 20 15 Frequency 10 5 0 Feb-08 Feb-09 Feb-10 Feb-11 Feb-12 Aug-07 Aug-08 Aug-09 Aug-10 Aug-11 Aug-12 Nov-07 Nov-08 Nov-09 Nov-10 Nov-11 Nov-12 May-08 May-09 May-10 May-11 May-12

Losing Narrative Winning Narrative Incomplete Narrative

The codebook for analyzing the media articles measured the article’s topical focus, presence/definition of policy problems and solutions, use of evidence, presence/type of characters, opinion on agricultural biotechnology, episodic/thematic focus, and types of risks/benefits. Two coders were trained to code the articles to attain a mutual understanding of the codebook. A standard set of instructions were established to foster intra- and intercoder reliability (Krippendorff, 2004). Following the training, intercoder reliability was tested on 10 percent of data randomly chosen from the population. Intercoder reliability was established wherein agreement ranged from 69 percent – 100 percent (π = .60 – 1.0)5. These intercoder measures were achieved with the following process: (1) coding of a random sample of articles,

(2) discussion by the coders and codebook revisions, and (3) recoding of a new set of randomly chosen 10 percent articles. The coded data were analyzed using Microsoft Excel and R Studio

5 Intercoder reliability scores are provided in Table 3.3. For Scott’s π results for coded data, a general guideline is to use data with a value of .80 and above, whereas exploratory and tentative conclusions can be drawn from moderately reliable data with Scott’s π ranging from 0.60 - 0.80 (Krippendorff, 2004).

39 statistical software package as appropriate. The categories for problem definitions and types of characters were analyzed qualitatively. Adhering to the ‘policy narrative’ definition (McBeth,

Jones, et al., 2014b), media articles without a policy referent (the main focus of the article was not on the Bt eggplant policy issue) and a character were removed from the final dataset leading to 171 articles (Times of India - 51 and Hindustan Times - 120).

Research Findings and Discussion

This study focused on the following narrative variables: policy problems and solutions, use of scientific evidence, discussion of risk and benefit information, and presence of characters.

Each narrative variable is discussed individually in context of whether the presence of a problem definition is related to other narrative variables in the media dataset. Narrative variables are simultaneously examined across losing, winning, and incomplete narratives in the media dataset.

This is followed by a broader discussion on the value of media articles as policy narratives with a focus on their narrativity score and the significance of a comparative analysis between two national level media.

Policy problems play an important role in a policy narrative since identifying a policy issue helps establish the plot. Policy solutions are also often provided in relation to a pre-defined problem enabling actors to narrow down the scope of a solution and providing agency to actors to solve the problem. Problem definitions usually include a discussion on the need for policy change or transformation (Kingdon, 2003; Shanahan et al., 2013; D. Stone, 2011). Based on the qualitative descriptions of problem definitions in the overall media dataset, seven categories were developed (see Table 3.2). Risks, health, and environmental impacts arising from the adoption of Bt eggplant were discussed most frequently (28.78 percent, n=40), followed by the problem of lack of adequate testing (20.14 percent, n=28), and socio-economic impacts (16.55

40 percent, n=23). This finding is supported in the literature since risks from agricultural biotechnology have been a point of contention among stakeholders and have contributed to a gridlock in negotiations. The problem definition related to lack of adequate testing facilities has also been fiercely debated (see e.g., Gupta, 2011).

A majority of policy narratives included a problem definition (81.29 percent, n=139) and the categories of definitions support the direction in which policy change is being suggested. For example, the frequent discussion of risk, health, and environmental impacts supports the minister’s proclamation that the moratorium would remain until studies prove: “the safety of the product from the point of view of its long-term impact on human health and environment”

(Bagla, 2010, p 767). This is supported through Figure 3.2, which shows that at the time of the policy decision, the problem definitions related to risk, health, and environmental impacts; and lack of adequate testing were most frequently discussed in the media.

41 Table 3. 2: Categories of Problem Definitions in Times of India and Hindustan Times

Problem Definition Percentage (n) Category 1: Data access issues 4.32 (6) 2: 28.78 (40) Risks/Health/Environment al Impacts 3: Socio-economic Impacts 16.55 (23) 4: Regulatory authority issues 11.51 (16) 5: Lack of adequate testing 20.14 (28) 6: Propaganda 7.91 (11) 7: Other 10.79 (15) Total: 100 (139)

Policy Problem Present 81.29 (139) Policy Problem Absent 18.71 (32) Total: 100 (171)

Figure 3. 2: Focus of Coverage over Time in Times of India and Hindustan Times

Focus of Problem Definitions 12 10 8 6 4 2 0 Feb-08 Feb-09 Feb-10 Feb-11 Feb-12 Aug-07 Aug-08 Aug-09 Aug-10 Aug-11 Aug-12 Nov-07 Nov-08 Nov-09 Nov-10 Nov-11 Nov-12 May-08 May-09 May-10 May-11 May-12

Data Access Issues Risks/Health/Envt Impacts Socio-economic Impacts Regulatory Authority Issues Lack Of Adequate Testing Propaganda Other

Table 3.3 examines the presence of problem definitions in relation to other narrative variables. Majority of the articles with a problem definition also included a solution. The presence of policy solutions and problems are statistically significant (p<0.001). A statistically

42 significant relationship was also found between the presence of risk and benefit information and the presence of a policy problem (p<0.001; Cramer’s V = 0.29). Therefore, narratives that define a policy problem also provide solutions and include risk and benefit information. No significant relationship was found between presence of a policy problem and the five characters.

Table 3. 3: Policy Problems and Presence of Evidence, Solutions, and Risk/Benefit Information in Times of India and Hindustan Times6

Policy Problem No Evidence % (n) Evidence Used % (n) Total % (n) Absence 22.92 (22) 13.33 (10) 18.71 (32) Presence 77.08 (74) 86.67 (65) 81.29 (139) Total 100.00 (96) 100.00 (75) 100.00 (171) X-squared7 = 1.9512, df = 1, ns Policy Problem No Solution % (n) Solution Used % (n) Total % (n) Presence 11.51 (16) 88.49 (123) 100.00 (139) X-squared = 171, df = 2, p-value < 0.0018 Policy Problem Absence of Risk/ Benefit Presence of Risk/ Total % (n) Information Benefit Information Absence 38.30 (18) 11.29 (14) 18.71 (32) Presence 61.70 (29) 88.71 (110) 81.29 (139) Total 100.00 (47) 100.00 (124) 100.00 (171) X-squared = 14.615, df = 1, p-value < 0.001; Cramer’s V = 0.29

Analyzing the use of narrative variables across losing, winning, and incomplete narratives, Table 3.4 shows how different problem definitions appeared across the narratives.

Risk, Health, and Environmental impacts were discussed most frequently by winning (40.68 percent, n=24) and incomplete (27.12 percent, n=16) narratives. But these did not appear in losing narratives. A closer qualitative examination shows that winning narratives discussed these

6 Intercoder reliability scores: 69.6% (π = .60) for types of risks/benefits expressed in the article, 78.3% (π = .67) for the use of evidence to support a policy preference, 91.3% (π = .84) for episodic/thematic focus, 91.3% (π = .87) for opinion of article on agricultural biotechnology, 95.7% (π = .91) for presence of policy solution to 100% (π = 1.0) for variables measuring presence of policy problem, and inclusion of characters 7 Pearson’s chi-squared test conducted with Yates continuity correction for all results in this article. 8 It is not possible to provide a measure of association (Cramer’s V) for these two elements since policy solutions were coded only for those narratives that had a policy problem. So, since policy problems showed no variation, a test of association is not possible.

43 impacts in relation to risks to health and environment. Socio-economic impacts were discussed more by the losing narratives (56.63 percent, n=10) and this was largely in context of how presently the yield is low and how it would benefit the economy and farmers to adopt Bt eggplant.

Table 3. 4: Problem Definitions Used by the Losing, Winning, and Incomplete Narratives in Times of India and Hindustan Times

Problem Definition Category (Final) Losing Winning Narrative Incomplete Narrative % (n) Narrative % (n) % (n) 1: Data access issues 5.26 (1) 1.69 (1) 6.78 (4) 2: Risks/Health/Environmental 0.00 (0) 40.68 (24) 27.12 (16) Impacts 3: Socio-economic Impacts 52.63 (10) 13.56 (8) 8.47 (5) 4: Regulatory authority issues 0.00 (0) 8.47 (5) 18.64 (11) 5: Lack of adequate testing 10.53 (2) 23.73 (14) 20.34 (12) 6: Propaganda 15.79 (3) 6.78 (4) 6.78 (4) 7: Other 15.79 (3) 5.08 (3) 11.86 (7) Total: 100.00 (19) 100.00 (59) 100.00 (59)

Table 3.5 shows a significant difference in the presence of policy problems across all three narratives in the media (X-squared = 8.0944, df = 2, p-value < 0.05, Cramer’s V = 0.22).

Policy problems occurred more frequently in the losing (86.36 percent, n=19), winning (90.77 percent, n=59), and incomplete (72.84 percent, n=59) narratives. There is no significant difference in the use of evidence (X-squared = 8.0944, df = 2, ns). However, losing narratives use evidence more frequently (63.64 percent, n=14) than winning (56.92 percent, n=37) and incomplete (61.73 percent, n=50) narratives. Thus, more frequent use of evidence may not be associated with narrative strategies to successfully influence policy outcomes. This supports similar work in the NPF (McBeth et al., 2007; Shanahan, McBeth, et al., 2011). All three narratives have a higher presence of policy solutions: losing (89.47 percent, n=17), winning

(89.83 percent, n=53), and incomplete (86.44 percent, n=51); and risk and benefit information:

44 losing (72.73 percent, n=16), winning (83.08 percent, n=54), and incomplete (65.43 percent, n=53); but no significant difference in their use. It should be noted that the p value < 0.06

(Cramer’s V = 0.18) for risk and benefit information, so we should not completely disregard the lack of significance. The discussion of risks and benefits of agricultural biotechnology is an important aspect of the policy debate with environmental and health related risks being frequently cited as problems.

45 Table 3. 5: Narrative Variables across Losing, Winning, and Incomplete Policy Narratives in Times of India and Hindustan Times

Losing Narrative Winning Narrative Incomplete Narrative Total % (n) % (n) % (n) % (n) Absence of Policy 13.64 (3) 9.23 (6) 27.16 (22) 18.45 (31) Problem Presence of Policy 86.36 (19) 90.77 (59) 72.84 (59) 81.55 (137) Problem 100.00 (22) 100.00 (65) 100.00 (81) 100.00 (168) X-squared = 8.0944, df = 2, p-value < 0.05; Cramer’s V = 0.22 Absence of 36.36 (8) 56.92 (37) 61.73 (50) 56.55 (95) Evidence Presence of 63.64 (14) 43.08 (28) 38.27 (31) 43.45 (73) Evidence 100.00 (22) 100.00 (65) 100.00 (81) 100.00 (168) X-squared = 4.5361, df = 2, ns Absence of Policy 10.53 (2) 10.17 (6) 13.56 (8) 11.68 (16) Solution Presence of Policy 89.47 (17) 89.83 (53) 86.44 (51) 88.32 (121) Solution 100.00 (19) 100.00 (59) 100.00 (59) 100.00 (137) X-squared = 0.35704, df = 2, ns Absence of 27.27 (6) 16.92 (11) 34.57 (28) 26.79 (45) Risk/Benefit Info Presence of 72.73 (16) 83.08 (54) 65.43 (53) 73.21 (123) Risk/Benefit Info 100.00 (22) 100.00 (65) 100.00 (81) 100.00 (168) X-squared = 5.7281, df = 2, p-value < 0.06, ns; Cramer’s V = 0.18

Next, the study examined appearance of characters in the overall media dataset as well as across the three narratives. In addition to heroes, villains, and victims, beneficiaries are included since actors supporting Bt eggplant have often invoked a discussion of who benefits from the crop. Also, certain character types were mentioned as potential or latent resources and were proposing or taking action but could not be categorized as hero, villain, victim, or beneficiary.

For example, Greenpeace sought access to data regarding testing of Bt eggplant (Raaj, 2008) or

Mahyco appealed to the High Court to not allow public release of data (Raaj, 2008). These were categorized as ‘Other.’ Use of characters is often related to the persuasiveness of a narrative

46 where characters may invoke sympathy from the audience and garner support (Jones, Shanahan, et al., 2014). This may be particularly prevalent in intractable policy issues where characters depict themselves as heroes to gather support for their policy preference and depict their opponents as villains.

47 Table 3. 6: Presence of Character Types in Times of India and Hindustan Times

Characters % (n)

Overall Heroes 80.59 (220) Villains 6.96 (19) Victims 3.30 (9) Beneficiary 2.93 (8) Other 6.23 (17) Total 100.00 (273)

Hero Business 2.73 (6) Government 49.09 (108) NGO/Environm 35.00 (77) ent Other 13.18 (29) Total 100.00 (220)

Villain Business 52.63 (10) Government 10.53 (2) NGO/Environm 21.05 (4) ent Other 15.79 (3) Total 100.00 (19)

Victim Environment 11.11 (1) Economy 11.11 (1) Other 77.78 (7) Total 100.00 (9)

As Table 3.6 shows, 80.59 percent (n=220) of the total characters identified were heroes.

This was surprising, given the extreme polarization in the policy debate. The expectation was to see a policy landscape with a high frequency of heroes and villains battling it out to achieve their desired policy preference. One reasoning for this highly unexpected finding is that characters were often advocating their solution to the policy problem without highlighting a villain. For

48 these heroes, agricultural biotechnology itself was a negative entity (but agricultural biotechnology was not portrayed as a villain per se) and hence, they did not feel the need to villainize others.

Previous NPF research shows that heroes are more likely to persuade an audience (Jones,

2013). Among the 220 hero characters in this study, the highest concentration was with the government. This is not surprising since three government ministers were highly active in the policy issue. This was followed by non-governmental organizations (NGOs) and the environment as heroes. Only six heroes were from the industry and 6.23 percent of all characters were classified as ‘Other.’ A closer qualitative analysis shows that these included actors mostly from NGOs and researchers. Of all characters, 2.93 percent were identified as beneficiaries (n=8) and were largely farmers. Table 3.7 reflects the use of characters across the three narratives in the media dataset. Interestingly, there is a significant difference in the use of beneficiaries

(Fisher’s Exact Test, p < 0.01; Cramer’s V = 0.34) with losing narratives having more beneficiaries (22.73 percent, n=5).

49 Table 3. 7: Use of Characters across Losing, Winning, and Incomplete Policy Narratives

Losing Narrative Winning Narrative Incomplete Narrative Total % (n) % (n) % (n) % (n) Absence of Hero 4.55 (1) 1.54 (1) 6.17 (5) 4.17 (7) Presence of Hero 95.45 (21) 98.46 (64) 93.83 (76) 95.83 (161) 100.00 (22) 100.00 (65) 100.00 (81) 100.00 (168) X-squared = 1.9487, df = 2, ns Absence of Villain 86.36 (19) 89.23 (58) 90.12 (73) 89.29 (150) Presence of Villain 13.64 (3) 10.77 (7) 9.88 (8) 10.71 (18) 100.00 (22) 100.00 (65) 100.00 (81) 100.00 (168) X-squared = 0.25599, df = 2, ns Absence of Victim 90.91 (20) 98.35 (61) 95.06 (77) 94.05 (158) Presence of Victim 9.09 (2) 6.15 (4) 4.94 (4) 5.95 (10) 100.00 (22) 100.00 (65) 100.00 (81) 100.00 (168) X-squared = 0.54063, df = 2, ns Absence of 77.27 (17) 100.00 (65) 96.30 (78) 95.24 (160) Beneficiary Presence of 22.73 (5) 0.00 (0) 3.70 (3) 4.76 (8) Beneficiary 100.00 (22) 100.00 (65) 100.00 (81) 100.00 (168) Fisher’s Exact Test, p <0.01; Cramer’s V = 0.34 Absence of Other 95.45 (21) 96.92 (63) 88.89 (72) 92.86 (156) Presence of Other 4.55 (1) 3.08 (2) 11.11 (9) 7.14 (12) 100.00 (22) 100.00 (65) 100.00 (81) 100.00 (168) X-squared = 3.767, df = 2, ns

Lastly, the study examined the differences in overall narrativity between the three types of narratives and between the two media sources: Times of India and Hindustan Times. Most of the narratives (see Table 3.8) only include half of the narrative components (mean = 4.94) with

38.01 percent (n=65) narratives having 5 and 26.32 percent (n=45) having 6 narrative components. Therefore, it is important to consider whether policy narratives generated in the media are producing the most effective narratives since they do not always include all the important and relevant components. Losing narratives used slightly more narrative components than winning; and winning narratives used slightly more components than incomplete ones.

Therefore, a higher use of narrative components may not always be connected to winning. One- way ANOVA reveals a statistically significant difference in the mean use of narrative

50 components (p<0.01) across the three types. There was a statistically significant difference in the mean use of narrative components (t-statistic = 3.105, p <0.002) between Times of India and

Hindustan Times. Since media coverage may vary across media outlets based on ideology, ownership, etc., it may be important to include multiple media sources to ensure that a particular ideology does not dominate the coverage.

51 Table 3. 8: Narrativity9 Index for Losing, Winning, and Incomplete narratives; and for Times of India and Hindustan Times

Number of Narrative % (n) Components 0 0.00 (0) 1 0.00 (0) 2 8.19 (14) 3 9.94 (17) 4 8.19 (14) 5 38.01 (65) 6 26.32 (45) 7 8.19 (14) 8 1.17 (2) 9 0.00 (0) 10 0.00 (0) Mean Narrativity Score Mean (n; SD) Full Dataset 4.94 (171; 1.39) Losing Narratives 5.45 (22; 1.44) Winning Narratives 5.17 (65; 1.23) Incomplete Narratives 4.63 (81; 1.44) p<0.01 (One-way ANOVA) Times of India 5.35 (51; 0.96) Hindustan Times 4.76 (120, 1.50) t-statistic = 3.105, p <0.002 Total 100.00 (171)

Conclusion and Future Directions NPF provides the theoretical and empirical tools to analyze the role of narratives in the policy process. One of the goals of this study was to expand scholarship on the NPF by exploring connections between narrative variables and policy outcomes in a non-U.S. context in order to test the generalizability of this policy process framework through its application in a policy subsystem in a different national and cultural context. Prior meso-level NPF research conducted in this context has revealed challenges from lack of data or accessibility issues for public consumption documents distributed by advocacy organizations (Gupta et al., 2014; Weible et al.,

9 Narrativity is a scale variable from 0-10, with 0 = no narrative components and 10 = used all narrative components (use of policy referent, policy problem, policy solution, hero, villain, victim, beneficiary, other, evidence, risks, and benefits).

52 2016). Hence, this study examined policy narratives available in the media to test the applicability of the NPF and found that narrative variables present across media provide substantial insight into the agricultural biotechnology policy issue in India. There was variation in articulation of problem definitions and characters among coalitions. Thus, in contexts where public consumption documents from advocacy organizations may not be available or easily accessible for meso-level NPF studies, media may prove to be an important and robust source of data. This study attempted to expand the scope of NPF research by using media narratives in a policy and national context not studied before.

In addition to the three traditional NPF characters (hero, villain, victim), this study included a new character type—beneficiary—previously used in Weible and others (2016). This is an important character type for certain policy issues, especially when characters are advocating for solutions benefiting a part of the population. Among the characters, a statistically significant relationship was found only in the use of beneficiaries across the three narratives. Also, potential and latent sources need to be considered in some form since they may be providing resources to advocate for a policy preference, for example, in the case of Greenpeace petitioning the legal system to release safety data. For characters, although no significant relationship was found in the use of heroes and policy problems in the media, given the large number of heroes, it may be important to consider how they appeared and if they articulated solutions.

Although risk perceptions have not been as widely studied in the NPF, risk perceptions may influence policy decisions that are aimed at risk reduction (Crow et al., 2016). In the present study, higher frequency of problem definitions related to risk, health, and environmental impacts support existing literature that these aspects were highly contested in the policy subsystem. The frequency of these problem definitions increased substantially at the time of the policy decision.

53 The higher frequency of risk-related concerns closely correlates with problem definition categories supporting the recommendation that risk information be included as a narrative component for policy issues fraught with risk. To test the generalizability of this component,

NPF research may benefit from further application of this component in risk-related areas such as science and technology issues and hazards and natural disasters. In case of evidence, this study found no significant difference in use of evidence, though it was used more by losing narratives. This supports similar work in the NPF (McBeth et al., 2007; Shanahan, Jones, et al.,

2011). In addition to the narrative core elements, additional components such as risk and evidence may need to be included based upon the policy issue under consideration.

Another goal of this study was to assess the role of incomplete policy narratives in policy debates and outcomes in an attempt to refine the definition of policy narratives and consider which narratives are important from empirical and audience reception perspectives. This study found that policy narratives do not always contain a full suite of narrative components since most of the narratives only included half of the narrative components. Recent research shows that the current NPF structure does not take into account several differences that exist among media organizations, platforms, and actors and it remains unclear whether media actively and voluntarily or incidentally participate in coalitions as part of their professional positions (Crow &

Lawlor, 2016). Media as conduits and contributor have been studied (Shanahan et al., 2008) but by including the category of incomplete policy narratives, one may account for an additional role of the media as disseminator of information without taking sides on a policy issue. This may also account for the differences among policy narratives distributed by advocacy organizations

(which have vested interests in promoting their policy preference), on the one hand, and by the media (which may not always promote a policy preference), on the other hand. Furthermore, the

54 two media sources that were examined show a significant difference in their use of narrative components. To avoid bias that may exist across media outlets and to ensure a robust dataset with breadth and depth of coverage, it may be useful to include multiple media sources. A useful future step might be to analyze how the use of specific narrative elements varied across the two sources. This may also help examine whether media sources vary in their role as conduits, contributors, or disseminators in the policy process. Lastly, the study focused on the role of narratives and how narratives are communicated through language. Despite the understanding that narratives are constructions of language, cultural context and nuances of language have yet to be explored by NPF studies. These language characteristics of narratives are important to understand in order to assess the stability of policy narratives when there are multiple dominant languages in a political system. This study analyzed media narratives in English language newspapers in a country with no official national language, but which uses English and Hindi for official purposes (Government of India n.d.). Given this, it is justifiable to use English newspapers, but it would be valuable to analyze media coverage of Hindi newspapers to examine variation in the use of narrative components and whether narrative construction varies across languages.

55 References

Arnold, R. D. 1990. The Logic of Congressional Action. New Haven: Yale University Press.

Bagla, Pallava. 2010. “After Acrimonious Debate, India Rejects GM Eggplant.” Science 327 (5967): 767. http://science.sciencemag.org/content/327/5967/767 (accessed April 23, 2018).

Bagla, Pallava, and Richard Stone. 2013. “Scientists Clash Swords over Future of GM Food Crops in India.” Science 340 (6132): 539-540. http://science.sciencemag.org/content/340/6132/539.full (accessed April 23, 2018).

Bandopadhyay, Rajib, Purnima Sinha, and Bratati Chaudhary. 2012. “Is Bt-brinjal ready for future food?--A critical study.” Indian Journal of Biotechnology 11 (2): 238–240.

Center for Environment and Education. 2010. National Consultations on Bt Brinjal: Primer on Concerns, Issues, and Prospects. http://www.ceeindia.org/cee/pdf_files/bt-brinjal- primer.pdf (accessed April 23, 2018).

Crow, Deserai A., and John Berggren. 2014. “Using the Narrative Policy Framework to Understand Stakeholder Strategy and Effectiveness: A Multi-Case Analysis.” In The Science of Stories: Applications of the Narrative Policy Framework in Public Policy Analysis (1st ed.), edited by M. D. Jones, E. A. Shanahan, and M. K. McBeth. 131-156. New York: Palgrave Macmillan Ltd.

Crow, Deserai. A., J. Berggren, L. A. Lawhon, E. A. Koebele, A. Kroepsch, and J. Huda. 2016. “Local Media Coverage of Wildfire Disasters: An Analysis of Problems and Solutions In Policy Narratives.” Environment and Planning C: Government and Policy 35 (5): http://journals.sagepub.com/doi/abs/10.1177/0263774X16667302 (accessed April 23, 2018).

Crow, Deserai A., and Andrea Lawlor. 2016. “Media in the Policy Process: Using Framing and Narratives to Understand Policy Influences.” Review of Policy Research, 33 (5): 472– 491. https://doi.org/10.1111/ropr.12187 (accessed April 23, 2018).

Damodaran, Harish, and Amitabh Sinha. 2016. “GM row again, with mustard topping.” The Indian Express, February 8. http://indianexpress.com/article/explained/gm-row-again- with-mustard-topping/ (accessed February 10, 2016).

Entman, Robert, M. 1995. “Public Opinion and the Media: How the Media Affect what People Think- And Think They Think.” In “Media” Res: Readings in Mass Media and American Politics, edited by J. P. Vermeer. 55-59. New York: McGraw-Hill.

Entman, R. M. (1995). Television, democratic theory and the visual construction of poverty. Research in Political Sociology, 7, 139–160.

56 Freelon, Deen G. 2010. “ReCal: Intercoder reliability calculation as a web service.” International Journal of Internet Science 5 (1): 20–33. http://www.ijis.net/ijis5_1/ijis5_1_freelon.pdf (accessed April 23, 2018).

Gupta, Kuhika, Joseph T. Ripberger, and Savannah Collins. 2014. “The Strategic Use of Policy Narratives: Jaitapur and the Politics of Siting a Nuclear Power Plant in India.” In The Science of Stories: Applications of the Narrative Policy Framework in Public Policy Analysis, edited by M. D. Jones, E. A. Shanahan, and M. K. McBeth. 89–106. New York: Palgrave Macmillan.

Government of India (n.d.). Constitutional Provisions: Official Language Related Part-17 of The Constitution of India. Department of Official Language, Government of India. http://rajbhasha.nic.in/en/constitutional-provisions (accessed March 22, 2017).

Herring, Ronald J. 2006. “Why did ‘operation cremate Monsanto’ fail? Science and Class in India’s Great Terminator-Technology Hoax.” Critical Asian Studies 38 (4): 467–93. https://doi.org/10.1080/14672710601073010 (accessed April 23, 2018).

Herring, Ronald. 2012. “State Science and Its Discontents: Why India’s Second Transgenic Crop Did not Follow the Path of Bt Cotton.” Weihenstephaner Socio-Economic Seminar, Center of Life and Food Sciences Weihenstephan, Technische Universität München (Vol. 13).

Herring, Ronald J. 2015. “State Science, Risk and Agricultural Biotechnology: Bt Cotton to Bt Brinjal in India.” Journal of Peasant Studies 42 (1): 159–186. https://doi.org/10.1080/03066150.2014.951835 (accessed April 23, 2018).

Indian Readership Survey. 2014. http://www.mruc.net/sites/default/files/IRS%202014%20Topline%20Findings_0.pdf (accessed April 24, 2018).

Iyengar, Shanto. 1990. “Framing Responsibility for Political Issues: The Case of Poverty.” Political Behavior 12 (1): 19–40. https://doi.org/10.1007/BF00992330 (accessed April 23, 2018).

Jayaraman, Killugudi. 2010. “Bt brinjal splits Indian cabinet.” Nature Biotechnology 28 (4): 296- 296. doi:10.1038/nbt0410-296 (accessed April 23, 2018).

Jones, Michael D. 2013. “Cultural Characters and Climate Change: How Heroes Shape Our Perception of Climate Science.” Social Science Quarterly 95 (1): 1–39. https://doi.org/10.1111/ssqu.12043 (accessed April 23, 2018).

Jones, Michael D., and Hank C. Jenkins‐Smith. 2009. “Trans‐Subsystem Dynamics: Policy Topography, Mass Opinion, and Policy Change.” Policy Studies Journal 37 (1): 37–58. https://doi.org/10.1111/j.1541-0072.2008.00294.x (accessed April 23, 2018).

57 Jones, M. D., and Mark K. McBeth. 2010. “A Narrative Policy Framework: Clear Enough to Be Wrong?” Policy Studies Journal 38 (2): 329–353. https://doi.org/10.1111/j.1541- 0072.2010.00364.x (accessed April 23, 2018).

Jones, Michael D., Elizabeth A. Shanahan, and Mark K. McBeth. 2014. The Science of Stories: Applications of the Narrative Policy Framework in Public Policy Analysis. Palgrave Macmillan.

Kingdon, John W. 2003. Agendas, Alternatives, and Public Policies. 2d edition. New York: Longman.

Krippendorf, Klaus. 2004. Content Analysis: An Introduction to Its Methodology. 2d edition. Thousand Oaks, CA: Sage Publications.

Lawton, R., and M. Rudd. 2014. “A Narrative Policy Approach to Environmental Conservation.” AMBIO, 1–9.

McBeth, Mark K., Michael D. Jones, and Elizabeth A. Shanahan.2014. “The Narrative Policy Framework.” In Theories of the Policy Process, 3rd ed., edited by Paul A. Sabatier and Christopher M. Weible. 225-265. Boulder, CO: Westview Press.

McBeth, M. K., E. A. Shanahan, P. L. Hathaway, L. E. Tigert, and L. J. Sampson. 2010. “Buffalo Tales: Interest Group Policy Stories in Greater Yellowstone.” Policy Sciences 43 (4): 391–409. https://doi.org/10.1007/s11077-010-9114-2 (accessed April 23, 2018).

McLeod, Douglas M., Gerald Kosicki, and J. M. McLeod. 2002. “Resurveying the boundaries of political communication effects.” In Media Effects: Advances in Theory and Research, edited by J. Bryant and D. Zillmann. 215–268. Mahwah, NJ: Lawrence Erlbaum Associates, Publishers.

Mohan, Vishwa. 2016. “Central regulator to take a final call on genetically modified Mustard on February 5.” Times of India, January 31. New Delhi. https://timesofindia.indiatimes.com/home/environment/Central-regulator-to-take-a-final- call-on-genetically-modified-Mustard-on-February-5/articleshow/50797661.cms (accessed February 10, 2016).

Nakyam, S. 2014. “Educational Decentralization Policy in Thailand: Unpacking Its Labyrinth to Pinpoint an Appropriately Further Step.” In International Conference on Public Administration. Hongkong, China. March 14-16, 2014.

Park, Young Sung. 2014. “A study of the construction permit process of 2nd Lotte World (skyscraper) using the Narrative Policy Framework.” The Korean Governance Review 21 (2): 101–125. http://www.dbpia.co.kr/Journal/ArticleDetail/NODE02474995 (accessed April 24, 2018).

Raaj, Neelam. 2008. “GM Brinjal battle goes to HC.” The Times of India, March 27. New Delhi.

58 https://timesofindia.indiatimes.com/india/GM-brinjal-battle-goes-to- HC/articleshow/2902812.cms (Accessed on May 2, 2018).

Ramaswami, B., and C. E. Pray 2007. “India: confronting the challenge–the potential of genetically modified crops for the poor.” In The Gene Revolution: GM Crops and Unequal Development, edited by Sakiko Fukuda Parr. 156–174. New York: Routledge.

Schlaufer, Caroline. 2016. “The Narrative Uses of Evidence.” Policy Studies Journal 46 (1): 90- 118. https://doi.org/10.1111/psj.12174 (accessed April 24, 2018).

Schattschneider, Elmer E. 1960. The Semi-Sovereign People. New York: Holt, Reinhart and Winston.

Shanahan, Elizabeth A., Mark K. McBeth, and Paul L. Hathaway. 2011. “Narrative Policy Framework: The Influence of Media Policy Narratives on Public Opinion.” Politics & Policy 39 (3): 373–400. https://doi.org/10.1111/j.1747-1346.2011.00295.x (accessed April 24, 2018).

Shanahan, Elizabeth A., Michael D. Jones, and Mark K. McBeth. 2011. “Policy Narratives and Policy Processes.” Policy Studies Journal 39 (3): 535–561. https://doi.org/10.1111/j.1541-0072.2011.00420.x (accessed April 24, 2018).

Shanahan, E. A., M. D. Jones, M. K. McBeth, and R. R. Lane. 2013. “An Angel on the Wind: How Heroic Policy Narratives Shape Policy Realities.” Policy Studies Journal 41 (3): 453–483. https://doi.org/10.1111/psj.12025 (accessed April 24, 2018).

Shanahan, E. A., M. K. McBeth, P. L. Hathaway, and R. J. Arnell. 2008. “Conduit or Contributor? The role of media in policy change theory.” Policy Sciences 41 (2): 115– 138. https://doi.org/10.1007/s11077-008-9058-y (accessed April 24, 2018).

Stone, Deborah A. 2011. Policy Paradox: The Art of Political Decision Making. London and New York: W.W. Norton New York.

United States Department of Agriculture (USDA). N.d. https://www.usda.gov/topics/biotechnology/biotechnology-frequently-asked-questions- faqs (accessed August 25, 2017).

Weible, C. M., K. L. Olofsson, D. P. Costie, J. M. Katz, and T. Heikkila. 2016. “Enhancing Precision and Clarity in the Study of Policy Narratives: An Analysis of Climate and Air Issues in Delhi, India.” Review of Policy Research 33 (2): 420-441. https://doi.org/10.1111/ropr.12181 (accessed April 24, 2018).

Weible, Christopher M., and E. Schlager. 2014. “Narrative Policy Framework: Contributions, Limitations, and Recommendations.” In The Science of Stories: Applications of the Narrative Policy Framework in Public Policy Analysis, edited by M. D. Jones, E. A. Shanahan, and M. K. McBeth. 235–246. New York: Palgrave Macmillan.

59 CHAPTER 4: POLICY NARRATIVES ACROSS TWO LANGUAGES: A COMPARATIVE STUDY USING THE NARRATIVE POLICY FRAMEWORK10

The Narrative Policy Framework (NPF) is based on the premise that stories are fundamental to human existence and that without stories communication would not be possible or at least be difficult. Given the importance of stories for individuals, it follows that stories would also be relevant for groups engaged in influencing the policy process to achieve their desired policy outcome. NPF studies these stories, which within the framework are referred to as policy narratives. It focuses on the role of these narratives, drawing from a rich scholarship in narrative, language, and culture. Distinguishing between structuralist and post-structuralist accounts of narratives, NPF is operationalized within the structuralist framework that asserts that there are “distinctive, generalizable narrative components such as characters and plots that exist across different contexts” (Jones, McBeth, & Shanahan, 2014, p. 4). On the other hand, the post- structuralist account of narrative considers the unique context of the narrative and its individual interpretation and asserts that each interpretation of a narrative varies based on the narrative and the individual interpreting the narrative. NPF refers to “the post-structural take on form and both the post-structural and structural takes on narrative content as the problem of narrative relativity”

(Jones, McBeth, & Shanahan, 2014, p. 5). Form refers to structure and content refers to the objects in the narrative (p. 4). NPF attempts to overcome narrative relativity of narrative form through “generalizable and context-independent narrative elements” (p. 5) mined from existing narrative scholarship; and addresses narrative relativity of narrative content “by empirically studying content in terms of strategy and belief systems” (Shanahan, Jones, McBeth, & Radaelli,

10 This chapter is a version of the following article and should be cited as follows: Huda, J. (2019). Policy Narratives across Two Languages: A Comparative Study using the Narrative Policy Framework. Review of Policy Research, 36(4), 523-546.

60 2018, p. 175). However, possible theoretical limitations of narrative form and content across languages remain an understudied area since NPF has not been applied to different linguistic contexts11.

An examination of policy narratives across different languages would allow researchers to examine the assumption that distinctive, generalizable narrative elements exist across multiple languages. Linguistics research shows that for cross-linguistic comparison one needs to create comparative concepts which would allow us “to identify comparable phenomena across languages and formulate cross-linguistic generalizations.” These comparative concepts need to be universally applicable (Haspelmath, 2010, p. 663). NPF research draws from a rich scholarship in narrative, language, and culture and has formulated such generalizable narrative components but these have not been tested in cross-linguistic comparison. This study mainly focuses on linguistic context, though policy narratives do not function in a vacuum and cultural context may also be relevant but is beyond the scope of this study.

NPF is a relatively new policy framework (Jones & McBeth, 2010) compared to other policy theories and frameworks but has been widely tested over the past decade in diverse policy contexts to further our understanding of the policy process. Its concepts are being more precisely defined (Merry, 2016; Schlaufer, 2016; Smith-Walter et al., 2016), new data sources are being used (Merry, 2016), and innovative methodologies (Gray & Jones, 2016; Weible et al., 2016) are being applied to study the NPF. NPF applications outside the United States reveal “the transportability of the NPF to diverse political systems and contextually nuanced policy domains” (Shanahan et al., 2018, p. 175). Thus, the groundwork for NPF has been laid in a diversity of contexts but the area of comparative linguistic analyses remains ripe for research.

11 Linguistic context refers to narrative construction in a particular language.

61 Comparative NPF research has been conducted in non-US policy contexts (Gupta et al., 2014;

Schlaufer, 2016; Weible et al., 2016) but has not explored narratives in languages other than

English or compared the same policy domain’s narratives across languages. Language characteristics of narratives in regional language narratives are important to understand in order to assess the stability of policy narratives when there are multiple dominant languages in a political system.

This study applies the NPF to agricultural biotechnology policy in India (Herring, 2015) and focuses specifically on narrative form in policy narratives characterized by narrative elements (setting, characters, plot, moral). It examines one of the core NPF assumptions that policy narratives have generalizable structural elements across multiple narrative contexts

(Shanahan et al., 2018). Given that any robust claim on generalizability would require extensive comparative NPF studies encompassing several languages, this study only takes a first step in that direction in the hope that future applications in other languages will help establish generalizability. With that aim in mind, it attempts to lend support to the hypothesis of transportability of NPF concepts not only into other contexts and settings which has been established (Shanahan et al., 2018, p. 199) but also into other languages to help move the NPF forward.

To do so, this study examines the commercialization process of a genetically modified

(GM) food crop, Bt eggplant, in India using NPF. This is an appropriate case study to use to apply the NPF across languages for two reasons. First, India is a linguistically diverse country with multiple official languages, which provides an excellent opportunity to study policy narratives across languages in the same policy subsystem. Second, agricultural biotechnology remains contentious despite its adoption almost three decades ago. Communication on

62 genetically modified organisms (GMOs) has been studied widely in terms of “how they are culturally received, socially constructed, and legitimized or delegitimized in public or media discourse” (Wenzelburger & König, 2017, p. 331). Research indicates that public opinion is influential in the regulation of GMOs (Tosun & Schaub, 2017). In India, intense public debate continues on this issue providing a rich dataset of policy narratives. More information on the specific case study is provided later. Research on framing in agricultural biotechnology indicates that supporters “frame the issue by eliciting emotions and impairing scientific evidence” and opponents emphasize scientific evidence (Tosun & Schaub, 2017, p. 311). Framing has been studied using the NPF (Crow & Lawlor, 2016; Shanahan et al., 2008) especially, in context of episodic and thematic framing (Iyengar, 1990). Although framing of agricultural biotechnology issues in India is a rich area to explore, this study focuses specifically on narrative elements sourced from NPF to examine transportability of these elements outside the English language.

The paper gives a brief overview of NPF (see Shanahan et al., 2018 for a broader overview) and its use in studying media narratives; and makes a case for studying policy narratives across languages. Background information on the case study is provided. Research methods are described along with research questions and their operationalization, followed by a focused description of each narrative element juxtaposed with expectations, findings, and discussion on the respective element; and a conclusion.

Narrative Policy Framework, Narrative Elements, and the Media

NPF is a policy process framework that examines policy narratives comprising narrative elements (narrative form) and analyzes the role of policy beliefs and strategies (narrative content) contained therein. NPF contends that policy narratives influence public opinion and the policy process. Its central questions revolve around the empirical role of policy narratives and

63 their influence on policy outcomes. NPF has four narrative core elements that comprise narrative form: setting (policy problems situated in a specific policy context), characters (heroes, villains, victims, beneficiaries, among others), plot (arc of action), and moral of the story (policy solution)

(Shanahan et al., 201812).

Narrative Elements

A policy narrative usually addresses a specific policy problem situated within a specific context called the setting. The policy problem helps establish the plot and can help narrow the scope of proposed solutions. Recognizing that a problem exists aids in acknowledging that it may be solved through policy making. Thus, when policy problems are defined, they are often accompanied by a policy solution (referred to as moral of the story). These are “the moral or normative actions incarnate” (Shanahan et al., 2018, p 176) and enable narratives to move beyond critique or argument (Jones, 2013). Actors in coalitions and policy entrepreneurs offer solutions as they work toward a policy goal and can help throw light on advocacy positions of actors (DCrow et al., 2017; M. D. Jones, McBeth, et al., 2014). Given the significance of policy problems and solutions, this study compares how problems and solutions are articulated across linguistic contexts over time to assess the stability of these components across languages.

Characters in policy narratives serve to persuade audiences and have traditionally been categorized as victims (harmed by the problem), villains (causing the problem), and heroes

(provide relief from harm/presume to have a solution) (Shanahan et al., 2018). But recent NPF research has explored additional character types including beneficiaries (Huda, 2018; Weible et al., 2016), allies and opponents (Merry, 2016), and entrepreneurs and charismatic experts

(Lawton & Rudd, 2014). Beneficiaries are an important character type, particularly when

12 This citation reflects the latest NPF theory but is built on previous NPF works. See (M. D. Jones & McBeth, 2010; McBeth, Jones, et al., 2014b; Shanahan et al., 2013; Shanahan, McBeth, et al., 2011).

64 characters advocate for solutions that benefit a portion of the population. Hence, beneficiaries are included in the study. The study also includes an additional character type ‘other,’ which refers to potential and latent sources that may provide resources to advocate for a policy solution but do not themselves fix the problem, cause it, are harmed by it, or benefit from a solution. For example, when government officials or scientists provide factual information without actively advocating for a solution. Although the narrative components described above are important in the construction of a policy narrative, these are unlikely to be present in all narratives. The most recent NPF scholarship continues to maintain that a policy narrative features a minimum of one character and a public policy referent but acknowledges that policy scholars can define narratives with different parameters (Shanahan et al., 2018). The present study adheres to the definition of a policy narrative cited here.

Although not typically treated as a distinctive narrative element, use of evidence is connected to the setting of a policy narrative. NPF research states that setting “consists of policy phenomena such as legal and constitutional parameters … evidence … or other features that some nontrivial amount of policy actors agree or assert are consequential within a particular policy area.” Evidence and other such features are treated more like props that “may become contested or the focal point of the policy narrative” (Shanahan et al., 2018, p. 176). NPF research on the use of evidence indicates that evidence may reflect diverging policy beliefs among coalitions (McBeth et al., 2010; Shanahan et al., 2013, 2008); may be used in support of solutions, though it is rarely provided (Shanahan et al., 2013, 2008); and may be used in different narrative strategies to indicate scientific uncertainty and disagreement (Gupta et al., 2014;

McBeth et al., 2007). Attempting to systematize the use of evidence in narratives, Schlaufer

(2016) has linked scientific evidence to all narrative elements. Accounting for the multi-

65 dimensionality of evidence use in policy narratives (Radaelli, Dunlop, & Fritsch, 2013), Smith-

Walter et al. (2016) further refine the use of evidence in NPF by distinctly categorizing evidence into scientific studies, statistics, polls, ipso dictum, and legal and argue that evidence is closely associated with other narrative elements. Within the policy arena of agricultural biotechnology, the authority of science has often been questioned and contradictory evidence abounds (Specter,

2014). Hence, use of evidence has been examined in this study in addition to the more typical core narrative elements outlined above.

Policy Narratives in the Media

Policy narratives studied within the NPF come from different sources and include media coverage (Crow et al., 2016; Shanahan, McBeth, et al., 2011; Shanahan et al., 2008), public consumption documents distributed by advocacy organizations (Gupta et al., 2014; Weible et al.,

2016), interviews (Gray & Jones, 2016; McMorris, Zanocco, & Jones, 2018), surveys (Jones,

2014), and social media (Merry, 2016). Media play an important role in shaping policy agendas

(Crow et al., 2016; Scheufele & Tewksbury, 2007) and help to understand how policy issues are brought to the attention of policymakers to bring about policy change (Baumgartner & Jones,

1993; Kingdon, 2003). Media may contribute to the policy agenda “by constructing (or co- constructing with policy advocates) the images used to communicate about and understand policy issues (Baumgartner & Jones, 2009); framing issues in certain ways

(Boykoff, 2011; Scheufele, 2000; Scheufele & Tewksbury, 2007); and disseminating the narratives communities use to discuss problems, policies, and solutions (Jones et al., 2014)”

(Crow et al., 2016, pp 5-6). NPF research has examined the role of media as conduits, contributors, or disseminators of policy narratives (Huda, 2018; Shanahan et al., 2008) and the media remain important, reliable, and accessible sources of policy narratives. While media have

66 changed radically over the past several decades, the traditional press including newspapers and magazines remain important and major sources of news and information for the public (Crow et al., 2016). The readership for regional media in India has only increased over the past several years. India has experienced a growth in the print media industry as compared to the West, which has witnessed declining print media circulations. The regional language press and television channels benefited immensely from investment by multinational corporations in the 1990s and expansion into new regions, which makes it relevant to analyze policy narratives in print media in India (Ninan, 2007; Rao, 2009).

Why Study Policy Narratives across Languages?

Exploring policy narratives in languages other than English provides an opportunity to explore nuances in language that may be present in policy narratives in other languages. News is a “representation of the world in language” and since language involves a semiotic code it imposes a structure on what it represents (Fowler, 2013, p. 4). This structure may vary based on language. Though the importance of language in news construction may be acknowledged, analysis of the structure itself may get ignored. NPF studies may benefit from examining the linguistic dimension of policy narratives to explore the power of language in the “social construction of reality” (p. 8). For example, NPF research claims that characters are central to narratives to persuade audiences, and if the audience can identify with and are sympathetic to a character, they are more likely to find that narrative persuasive (Jones, Shanahan, et al., 2014). In

NPF research conducted within the US cultural context, heroes have been found to be most likely to influence opinion (Jones, 2014). Other studies demonstrate a ‘devil shift’ wherein coalitions portray opponents as more evil and powerful and this narrative strategy is found to exert more influence on public opinion (McBeth et al., 2012; Sabatier, Hunter, & McLaughlin, 1987). NPF

67 research in other contexts may reveal differences in the preponderance of character types and narrative components prevalent in policy narratives.

A small but growing number of NPF studies have been conducted in non-US contexts as mentioned earlier. But these content analyses have largely remained confined to the English language. Although English language media predominate in the US context and some other regions of the developed world, this does not hold true for the developing or less developed regions. For example, in India 90% of newspaper readership comprises regional language media with only one English language newspaper, Times of India, ranked eleventh, in the top twenty most-read newspapers in the country (IRS, 2017). India is a country with no official national language but uses English and Hindi for official purposes (GoI, n.d.). English language media are largely consumed by policy elites and provide an important glimpse into the narratives of stakeholders. However, a singular focus on English language media would likely result in a marginalization of portrayals outside of the English language media, which are routinely consumed by certain stakeholders and India’s non-urban public.

Regional language media13 analysis may reveal how local language media differs in the way information is delivered and what information is delivered. It can help examine if there are different, more nuanced concerns highlighted or may reveal potentially diverging issue foci.

Regional language media are expected to focus on local as opposed to global stories. India’s top two national Hindi newspapers employ local journalists. This localization has enabled news to travel closer to the readers and increase the presence of Hindi media in rural areas. Through a focus on local issues, Hindi media has placed regional issues alongside national issues thereby

13 Regional language media refers to media disseminated in the regional language. Its audience may be national, regional, or local depending upon the coverage of the newspaper. This study examines a national level regional language newspaper.

68 creating an alternative public discourse (Friedlander, Jeffrey, & Seth, 2001; Neyazi, 2010, 2011).

Also, Hindi newspapers are sensitive to “local cultural values” and can reproduce the “global context locally” such as repackaging photos of the singer, Madonna, to align with local preferences (Neyazi, 2010, pp 907-908).

Given the different social groups that purchase Hindi and English newspapers, “one can say things in English-language newspapers that one cannot say in Hindi newspapers”

(Friedlander et al., 2001, pp 160-61). Editors may sanitize or change content to suit cultural or religious preferences of their readers, which may vary drastically across a country as large as

India. Audience for Hindi and English media may differ in their issue interest because elites are more likely to consume English media and Hindi readers are more likely to focus on issues of daily or local importance (Sahay, 2006). Thus, Hindi media content may be more region specific even when reported in a national newspaper. By examining not only what is present but also what is absent or silenced in different language media, such regional language media analyses may reveal marginalization; may provide insights into locally important issues and their interaction with issues in the national or global discourse found in the English media; or may reveal linguistic limitations in conveying technical topics (Ghosh & Boykoff, 2018; Neyazi,

2010). Variation in coverage may influence composition of policy narratives and diverging foci of coverage may influence frequency of narrative elements.

India is a linguistically diverse country with over a dozen official languages. English media has dominated the national media market but since 1990s there has been a boom in the

Hindi news media industry (Neyazi, 2011). Hindi is spoken by around 41.03% of the population

(Jainl, 2014). Dainik Jagran, India’s most widely read Hindi newspaper has a circulation of over seventy million copies (IRS 2017, n.d.). Hindi news media are not only prevalent in north India

69 where Hindi is widely used but have spread to other regions (Neyazi, 2011). The proliferation of

Hindi media has enabled wider public participation of marginalized groups who could not participate in the policy process dominated by the English-speaking elite. This makes it salient to use Hindi media in this study.

There may be ideological differences among Hindi and English media that may influence the composition of narratives. Although these ideological differences are not explicitly analyzed in this study, it is important to acknowledge their possible existence and potential role in any variation that may exist in policy narratives. As will be discussed in the Methods section below, the dataset for the Hindi media coverage is comparatively smaller than the English media dataset. The focus on local issues in Hindi newspapers in spite of having a national coverage may be one potential reason for the smaller Hindi media dataset that was gathered for this study as the policy issue under consideration may not have been as prevalent at the local level.

Comparing policy narratives in Hindi and English media, this study applies the Narrative

Policy Framework to the issue of agricultural biotechnology policy in India. Background information on the case is below.

Background Information: Bt Eggplant Commercialization Process

Humans have been manipulating the genetic make-up of plants and animals since the beginning of agriculture over 10,000 years ago through selectively breeding the best adapted species. Through the application of genetic principles, conventionally or through the use of biotechnology, they introduce desirable traits into crops or livestock. According to the

Convention on Biological Diversity (CBD), biotechnology is defined as “any technological application that uses biological systems, living organisms, or derivatives thereof, to make or

70 modify products for specific use” (Secretariat of the Convention on Biological Diversity 1992, n.p.).

Agricultural biotechnology was first adopted in the United States in 1996 and remains under contention globally (Herring, 2006, 2015). India is an excellent country to study stakeholder attitudes to and policy surrounding GM crops since it is particularly vulnerable to food security issues, and agricultural biotechnology has often been promoted as a promising means of ensuring food security. The issue remains contentious with a plurality of perspectives ranging from health/environmental risks to Intellectual Property Rights issues and the threat of monopolization by international corporations like Monsanto.

India has a long history of promoting agricultural biotechnology since mid-1980s. India’s first GM crop, Bt cotton, was adopted in 2002. Despite the widespread cultivation of Bt cotton in

India, other crops have not received approval for commercial cultivation. In mid-2000, Bt eggplant14 or Bt brinjal, as is commonly known in India, was perceived as the first genetically modified food crop to be considered seriously by the regulatory system (Ramaswami & Pray,

2007). Eggplant is a popular vegetable crop in India among small scale farmers as well as among low income consumers. After nine years of testing by the regulatory body of genetic engineering technology, Genetic Engineering Approval Committee (GEAC), Bt eggplant was approved for growing in India in October 2009 (Bagla & Stone, 2013; Herring, 2012, 2015). However, in

2010, Bt eggplant could no longer be planted legally after the Federal Minister for Environment and Forests rejected GEAC’s decision and placed an indefinite moratorium pending further evaluations citing strong public opposition (Bandopadhyay et al., 2012; Laursen, 2012; Samuels,

14 This transgenic eggplant is created through the insertion of a gene cry1Ac from the soil bacterium (Bacillus thuringiensis) into eggplant and is said to provide the plant with resistance against lepidopteran insects like the Brinjal Fruit and Shoot Borer (Leucinodes orbonalis) and Fruit Borer (Helicoverpa armigera) (Bandopadhya, Sinha, & Choudhary 2012).

71 2013). The moratorium overturned the regulatory bodies’ recommendation to allow the commercialization of Bt eggplant. This study analyzes policy narratives surrounding the policy process of the commercialization of Bt eggplant in India.

Methods

Data for the study were collected from the leading national Hindi (Dainik Jagran) and

English (Times of India and Hindustan Times) newspapers in India. The data collection and analysis for this research took place in four phases. In the first phase, policy documents were gathered for contextual policy information such as timeline, actors involved, and policy issue information. In the second phase, data from the English newspapers (n = 171) were collected and analyzed using an NPF focused codebook following NPF coding procedures15 (Huda, 2018). The

NPF codebook has become stable over time and many scholars have applied it to varied policy domains (see Shanahan, Jones, & McBeth, 2018 for detailed discussion on NPF coding procedures). The sampling timeframe was three years before and after the decision to ban Bt eggplant: February 9, 2007 to February 9, 2013. Online archives of the New Delhi edition of the

English newspapers are available in the ProQuest News and Newspapers database.

In the third phase, archives of the New Delhi edition of the Hindi newspaper, Dainik

Jagran, were accessed in person in June-July 2017 at their corporate office in Noida, India. Due to accessibility issues, it was only possible to collect data from a single Hindi newspaper. Online archives of Dainik Jagran are not available. After a manual search of the dataset from the same

15 A total of 1,212 articles (397 from Times of India and 815 from Hindustan Times) were downloaded using search terms: “Bt brinjal” OR “Bt eggplant” OR “genetically modified” OR “agricultural biotechnology.” Duplicate articles or those not focusing on Bt brinjal or not written in narrative form were removed from the dataset leading to a final dataset of 227 articles (87 from Times of India and 140 from Hindustan Times) that were coded initially. Based on the ‘policy narrative’ definition (Shanahan, Jones, McBeth, et al., 2018), media articles without a policy referent and a character were removed leading to 171 articles (Times of India - 51 and Hindustan Times - 120) that were included in the initial study in the second phase.

72 sampling timeframe as the English media coverage, articles focusing on Bt eggplant (बीट$ बग& न) were collected (n=20). A preliminary review was conducted to ensure the articles adhered to the

‘policy narrative’ definition. One article was found to not be telling a story but merely enumerating a list of facts. This was removed from the final dataset leading to 19 articles from the Hindi media coverage. Media coverage from 2012 was not available at the corporate office and was reported missing.

Since the sample size for the Hindi media coverage (n=19) is relatively small compared to the English media coverage (n=171), a subset from the English media coverage has been used so as not to skew the results based on the relatively higher quantity of English media coverage.

The subset of English media coverage was constructed after a preliminary review of the events that were covered in the Hindi media dataset. Corresponding articles on similar events within a week (before/after) were located in the English media dataset. In order to create the subset, a list of events covered in each of the 19 Hindi media articles was created. For example, a Hindi article dated 17 Aug. 2007 covered the issue of approval given by the regulatory body for field trials of Bt eggplant, then articles within a week (before/after) of that date were searched in the

English dataset. If the same issue was covered by English media in that period, these were subsequently included in the English dataset used for this study. This led to a final subset of 52 articles from the English media dataset (14 from Times of India and 38 from Hindustan Times).

The search terms, daily circulation, and article counts used in the present analysis are included in

Table 4.1.

73 Table 4. 1. Newspapers, Search Terms, and Article Counts

Newspaper Language Audience Circulation Search Terms Article Counts English (Times of National 19,894,000 daily Bt brinjal, Bt eggplant, genetically 52 India and Hindustan modified, agricultural biotechnology Times) Hindi (Dainik Jagran) National 70,377,000 daily 19 Bt eggplant (बीट$ बग& न)

To illustrate the distribution of media coverage in the two datasets, Figure 4.1 shows media coverage of the Bt eggplant issue over the six-year timeframe only focusing on similar events covered by the Hindi (n=19) and English (n=52) media. As expected, there was relatively less coverage on the issue prior to the regulatory approval (Hindi=3, English=2) than after it. The

Hindi media coverage peaked just prior to the announcement of the ban on Bt eggplant imposed by the Federal Minister for Environment and Forests on 9 Feb. 2010 (Hindi=9), as the issue rose on the agenda, while English media coverage also peaked just prior to the ban but is more frequent following the ban wherein consequences of the ban were debated (English=35).

74 Figure 4. 1. Issue Coverage on Similar Events across Hindi (n=19) and English Media (n=52)

Coverage of the issue on similar events in Hindi (Dainik Jagran) and English (Times of India and Hindustan Times) media 35

30 Ban by Federal Minister After GEAC approval/before on Feb. 9, 2010 25 ban: Hindi: 9 English: 15 20

15 Before GEAC approval: After ban: Hindi: 3 10 Hindi: 7 English: 2 English: 35 5

0 Jun-08 Jun-09 Jun-10 Jun-11 Oct-07 Oct-08 Oct-09 Oct-10 Apr-08 Apr-09 Apr-10 Apr-11 Feb-08 Feb-09 Feb-10 Feb-11 Dec-07 Dec-08 Dec-09 Dec-10 Aug-07 Aug-08 Aug-09 Aug-10 Aug-11

Article Count Hindi Article Count English

The accepted standard for testing intercoder reliability is 10% but since the Hindi dataset was relatively small, 21.05% of the Hindi articles were tested for reliability to account for the small sample size in conjunction with the 10% that had been previously coded from the original larger English media dataset. Intercoder reliability scores are provided in Table 4.2.

Table 4. 2. Intercoder Reliability Scores

Narrative element Percentage agreement; Scott’s pi Use of Evidence 77.8%; 0.67 Presence of Policy Solution 96.3%; 0.92 Presence of Policy Problem 100%; 1 Inclusion of Characters 100%; 1

NPF primarily draws on quantitative methods but a small number of qualitative NPF studies have been conducted (Gray & Jones, 2016; McMorris et al., 2018; Pierce, Smith-Walter,

& Peterson, 2014). The fourth phase of data collection used in this study combined the traditional deductive analysis of policy narrative elements with inductive methods common to

75 qualitative analysis to further delve into what kind of policy problems and solutions were discussed across the two media, who were the policy actors involved, and what types of evidence were used. Such an in-depth qualitative analysis allows an exploration of how narrative elements across the two languages vary in highlighting specific problems and solutions, the types of actors that are more persuasive in one language over another, and the kind of evidence they use.

Qualitative approaches are especially advocated when exploring new areas and, hence, a qualitative analysis is appropriate for a novel cross-linguistic NPF study. Categories for problems and solutions were informed by an inductive approach drawing from grounded theory that identified categories that emerged during coding and analysis and leveraged concepts encountered in the literature review on agricultural biotechnology (Miles & Huberman, 1994).

Research Questions and Operationalization

To test the hypothesis of the transportability of narrative elements in multiple narrative contexts, this study is guided by the following research questions:

RQ1: Are narrative elements transportable outside of the English language?

RQ1a: If yes, are narrative elements (policy problems, policy solutions, characters,

evidence) used similarly in policy narratives in different languages on the same policy

issue?

To examine these questions, the study first traces whether the narrative elements under consideration do actually appear in Hindi narratives to test for transportability. To analyze whether narrative elements are used similarly in different languages, the study goes on to compare the use of policy problems, solutions, characters, and evidence in Hindi and English policy narratives. Narrative elements (problems, solutions, characters, and evidence) perform certain functions as per the Narrative Policy Framework. For example, it is usually heroes who

76 offer solutions, or villains who are blamed for problems. But research is lacking on whether these narrative elements function similarly in non-English languages. Research findings will first be presented below on whether narrative elements are transportable outside of the English language through an analysis of whether each element exists in Hindi narratives. Based on the literature, expectations will then be outlined for each narrative element, followed by a comparative analysis of that narrative element across Hindi and English narratives to examine narrative similarity across languages.

Expectations, Research Findings, and Discussion

An analysis of whether narrative elements exist in Hindi narratives (narrative transportability) is provided first, followed by a discussion on expectations for each element and a comparative analysis of that element between Hindi and English narratives (narrative similarity across languages).

RQ1: Are narrative elements transportable outside of the English language?

To examine the transportability of narrative elements, this study examined the appearance of policy problems, solutions, characters, and evidence across Hindi and English narratives.

Findings support the transportability of narrative elements to Hindi since all the elements under consideration appear in the Hindi dataset including the characters categorized in NPF research

(Table 4.3). These narrative elements also appear in similar frequency across the two datasets.

77 Table 4. 3. Narrative Elements Across Hindi and English Media

Hindi English Percentage (n) Percentage (n) Absence of Policy Problem 26.32 (5) 26.92 (14) Presence of Policy Problem 73.68 (14) 73.08 (38) Total 100.00 (19) 100.00 (52) Absence of Policy Solution 14.29 (2) 10.53 (4) Presence of Policy Solution 85.71 (12) 89.47 (34) Total 100.00 (14) 100.00 (38) Heroes 55 (22) 81.94 (59) Villains 15 (6) 6.94 (5) Victims 10 (4) 2.78 (2) Beneficiaries 2.5 (1) 1.39 (1) Others 17.5 (7) 6.94 (5) Total 100 (40) 100 (72) Absence of Evidence 57.89 (11) 57.69 (30) Presence of Evidence 42.11 (8) 42.31 (22) Total 100.00 (19) 100.00 (52)

RQ1a: If yes, are narrative elements (policy problems, policy solutions, characters,

evidence) used similarly in different languages on the same policy issue?

To examine whether narrative elements are used similarly across two languages on the same policy issue, the study analyzes whether there is any variation in terms of how the narrative elements appeared over time across the two media datasets. Expectations for how each narrative element has traditionally been used in NPF are discussed first followed by findings and discussion. Also, to better understand how narrative elements appeared over time in the policy process, the narratives were analyzed across three time periods over six years:

1. before GEAC approval – Feb. 9, 2007 – Oct. 13, 2009 (i.e. before the regulatory body,

Genetic Engineering Approval Committee (GEAC) recommended Bt eggplant for

commercialization)

78 2. after GEAC approval/before ban: Oct. 14, 2009 – Feb. 8, 2010 (i.e. after the regulatory

recommendation but before the ban was imposed by the Federal Minister for

Environment and Forests)

3. after ban: Feb. 9, 2010 – Feb. 9, 2013 (i.e. after the ban was imposed by the Minister for

Environment and Forests)

Policy Problems

A policy narrative is usually built around some stated policy problem (Shanahan et al.,

2013). Identification of the problem helps establish the plot and setting, and definition of the problem can narrow the scope of the proposed solutions. Recognizing that a problem exists allows for the acknowledgement that the problem could actually be solved through policy making and often calls for policy change or transformation. Stone (2011) observes that when a policy problem is defined, one needs to also look at “how that definition defines interested parties and stakes, how it allocates the roles of bully and underdog, and how a different definition would change power relations” (p. 247). Given this, how the problem is defined, who defines it, and its timing are important to consider. In the context of agricultural biotechnology policy, we would expect a wide range of problems being defined from health and environmental impacts to socio-economic impacts. But which problems are more salient may vary based on the timeline of the policy process.

Policy Problems – Findings and Discussion: Table 4.4 reports policy problems as they appeared over time across Hindi and English media coverage. In the Hindi media, policy problems were reported more frequently after GEAC approval and before the ban (n=8), whereas in the English media the frequency of policy problems was much higher after the ban (n=29). This indicates that the Hindi media highlighted issues with the GM crop while the policy was being debated in

79 the public domain whereas the English media discussed problems after the ban was imposed highlighting the continued need for the ban.

Table 4. 4. Appearance of Policy Problems over Time across Hindi and English Media

Percentage (n) Percentage (n)

Policy Problem

Hindi (n=19) English (n=52) Before GEAC Approval: 5.26 (1) 0.00 (0) After GEAC Approval and Before Ban: 42.11 (8) 17.31 (9) After Ban: 26.32 (5) 55.77 (29) Policy Problem Absent (overall) 26.32 (5) 26.92 (14) Total 100.00 (19) 100.00 (52)

Policy problems about Bt eggplant were categorized based on their foci: data access issues (related to GM testing data not being available in the public domain), risk/health/environmental impacts of the crop, socio-economic impacts of the crop, regulatory authority issues, lack of adequate testing of the crop, and ‘other’. When these problem definitions were examined over time (Figure 4.2), some variation is prevalent: English and Hindi narratives did not define a problem frequently before GEAC approval (n=2 for both) while one

Hindi narrative highlighted the lack of adequate testing of GM crops. After GEAC approval and before the ban, Hindi narratives discussed risk/health/environmental impacts (n=3), socio- economic impacts (n=1), and regulatory authority issues (n=1), whereas discussion on risk/health/environmental impacts was absent in the English narratives on those events during this period. These impacts were discussed in the English narratives only after the ban (n=6), although both Hindi (n=3) and English (n=9) narratives during this time focused relatively more on the issue of lack of adequate testing. Data access issues were not discussed at all in the Hindi narratives, while these were discussed in the English narratives after the ban (n=3). Thus,

80 discussion in the Hindi media progressed from issues with GM crops themselves (such as risk/health/environmental impacts) to issues with the policy process (such as regulatory authority issues) while the discussion in the English media focused more on the policy process than on issues with the GM crop.

81 Figure 4. 2. Focus of Problem Definitions over time

• Did not define a problem (2) • Lack of adequate testing Hindi (1)

Before GEAC Approval

• Did not define a problem (2) English

• Risk/Health/Environmental impacts (3) • Other (2) Hindi • Socio-economic impacts (1)

• Regulatory authority issues (1)

After GEAC Approval and Before Ban

• Did not define a problem (6) • Regulatory authority issues English (3) • Socio-economic impacts (2) •

• Lack of adequate testing (3) • Did not define a problem (2) Hindi • Regulatory authority issues (1)

After Ban

• Lack of adequate testing (9) • Other (8)

• Risk/Health/Environmental English impacts (6) • Did not define a problem (6) • Data access issues (3)

82 Although the lack of adequate testing of GM crops was a concern discussed in the Hindi narratives initially, the English narratives did not discuss it at the time. Concerns raised in the

Hindi narratives after GEAC approval and before ban included the risk/health/environmental impacts, socio-economic impacts, and regulatory authority issues. But surprisingly, risk/health/environmental impacts were not discussed in the English narratives at that time but were only discussed after the ban. When the ban was imposed, the Minister for Environment actually referred to these impacts and these may have been then picked up in the English media.

While the discussion in the Hindi narratives was more focused on issues related with GM crops themselves, the English narratives reflected a concern with the policy process itself since data access issues and lack of adequate testing of GM crops were discussed in the English media more than the impacts of GM crops. Thus, as expected we see a wide range of problems being defined but we also see substantial variation in how Hindi and English narratives defined problems over time reflecting different concerns in the two media. Although both datasets had a similarly high frequency of narratives highlighting problems, studying the narratives across the two datasets over a substantial period of time shows that problems were defined very differently across the two datasets.

Policy Solutions

A policy solution may accompany a policy problem defined in the narrative. It is usually a normative prescription and focuses character’s actions and motives (Shanahan, Jones, McBeth, et al., 2018). It provides guidance and direction for mobilization (McBeth et al., 2012). A solution that aims to control the policy outcome would enable a narrative to move beyond critique or argument (Jones, 2013). Policy solutions are not always present in a narrative and such narratives may instead highlight uncertainty (Jones, Shanahan, et al., 2014). Policy

83 solutions can help scholars understand advocacy positions since it is usually advocacy coalitions that offer solutions directed toward a specific policy goal. Solutions need not be supported by empirical or scientific evidence since the policy issue can go beyond the scientific aspect to include social, cultural, and political elements (McBeth, Lybecker, & Husmann, 2014). NPF research shows that presence of fewer solutions corresponds with a preponderance of victims and villains as it is heroes who typically offer solutions (Jones, 2013; McBeth et al., 2012). Given that solutions tend to provide direction for mobilization, the types of solutions may vary at different stages of the policy process. We would expect those opposing Bt eggplant to highlight more solutions calling for a ban, especially after GEAC approval whereas after the ban we would expect those supporting Bt eggplant to highlight solutions directed toward commercialization.

Table 4. 5. Appearance of Policy Solutions over Time across Hindi and English Media

Percentage (n) Percentage (n)

Policy Solution

Hindi (n=14) English (n=38) Before GEAC Approval: 7.14 (1) 0.00 (0) After GEAC Approval and Before Ban: 57.14 (8) 18.42 (7) After Ban: 21.43 (3) 71.05 (27) Problem Defined but No Solution (overall) 14.29 (2) 10.53 (4) Total 100.00 (14) 100.00 (38)

Policy Solutions – Findings and Discussion: The appearance of problems corresponds with solutions. Solutions appeared more frequently in the Hindi media after GEAC approval and before the ban (n=8), whereas in the English media solutions spiked after the ban (n=27) (Table

4.5). No solutions were prevalent in the English media before GEAC approval which aligns with the absence of policy problems at the time. Policy solutions were categorized according to: need for an independent regulatory authority; need for new independent studies/research; importance

84 of public/stakeholder engagement; need for legal action against GM developers; need for imposition of ban; role of the regulatory process; role of science, ethics, transparency; need to consider food security; and urging of commercialization of the GM crop. Solutions focusing on the need for new independent studies/research were more frequent than other types of solutions across both Hindi (n=4) and English (n=15) narratives. When solutions were compared across the three time periods (Table 4.6), English narratives did not focus on solutions at all before

GEAC approval (no problems were defined at the time in English media). The need for new independent studies/research were frequently discussed by both Hindi (n=2) and English (n=4) narratives after GEAC approval and before ban. Surprisingly, the actual imposition of the ban or commercialization of the GM crop were not frequently discussed by either media.

Commercialization was offered only once as a solution in the Hindi narrative while imposing the ban was suggested once in each of the narratives during different time periods. Solutions focusing on public and stakeholder engagement appeared in English narratives but were absent in Hindi narratives. Similarly, solutions focusing on the role of science, ethics, and transparency appeared in English narratives but were noticeably absent in Hindi narratives. Solutions focusing on the need for an independent regulatory authority were discussed in Hindi (n=2) and English

(n=4) narratives though in different frequencies across the time periods.

85 Table 4. 6. Focus of Policy Solutions over time

• Need for new • Need for new independent • Problem defined but independent studies/research (2) no solution offered (2) studies/research (1) • Role of the Regulatory • Independent Regulatory Process (2) Authority (1) • Independent Regulatory • Need for new Authority (1) independent • Imposition of Ban (1) studies/research (1) • Food Security (1) • Legal Action Against • Commercialization (1) GM Developers (1)

Hindi Media

BEFORE GEAC AFTER GEAC APPROVAL AFTER BAN APPROVAL AND BEFORE BAN

English Media

• Need for new independent • Need for new studies/research (4) independent • Problem defined but no studies/research (11) solution offered (2) • Independent • Public/Stakeholder Regulatory Authority Engagement (1) (4) • Role of the Regulatory • Public/Stakeholder Process (1) Engagement (3) • Role of the Regulatory Process (3) • Role of science, ethics, transparency (3) • Problem defined but no solution offered (2) • Legal Action Against GM Developers (2) • Imposition of Ban (1)

86 Although similar solutions appeared across the two media, these were offered variably over time and some solutions such as public/stakeholder engagement and role of science, ethics, transparency appeared in one dataset but not the other indicating that these aspects may not be as relevant to the regional language reader. Since solutions can direct mobilization efforts, focus on solutions asking for new independent studies/research would help advocate for a ban until new studies are conducted. This also aligns with a higher frequency of narratives opposing Bt eggplant after regulatory approval. After the ban, we would expect those supporting Bt eggplant to highlight solutions directed toward commercialization. We instead find that the focus stays on solutions asking for new independent studies/research as those against Bt eggplant continue to mobilize to uphold the moratorium.

Characters

Characters are categorized as victims (harmed by the problem), villains (causing the problem), heroes (provide relief from harm/presume to have a solution), beneficiaries (benefiting from the solution), and other (latent sources that provide resources but do not fix the policy problem, cause it, are harmed by it, or benefit from a solution). Characters often serve to persuade audiences. For example, if the audience identifies with a character or is sympathetic to it, they may find the narrative more persuasive (Jones, Shanahan, et al., 2014). Previous research shows that the hero is likely to be more persuasive in influencing opinions (Jones, 2014) and, in particular, local actors as heroes are found to be more persuasive (McBeth et al., 2012). Villains usually appear when coalitions frame opponents as evil and powerful (McBeth et al., 2012).

Given the persuasiveness of heroes and the portrayal of opponents as villains, one would expect to see a policy landscape with a high frequency of heroes advocating solutions and opponents cast as villains battling it out to achieve their desired policy preference.

87 Table 4. 7. Appearance of Characters over Time across Hindi and English Media

Percentage (n) Percentage (n)

Characters16 (Overall)

Hindi (n=40) English (n=72) Before GEAC Approval: 7.50 (3) 4.17 (3) After GEAC Approval and Before Ban: 55.00 (22) 41.67 (30) After Ban: 37.50(15) 54.17 (39) Total 100.00 (40) 100.00 (72)

Characters – Findings and Discussion: Frequency of characters increased after GEAC approval and before the ban in the Hindi media (n=22) while more characters appeared after the ban in

English media (n=39) (Table 4.7). Heroes were more frequent across both Hindi and English narratives. But their frequency varied across the three time periods (Figure 4.3). More heroes appeared in Hindi narratives after GEAC approval and before the ban (n=12) while in the

English narratives’ heroes appeared more frequently after the ban (n=35) compared to the other two time periods. In both narratives, beneficiaries only appeared after GEAC approval and before ban (n=1 for both) and victims (Hindi=4, English=2) and villains (Hindi=6, English=5) appeared only after GEAC approval.

16 n is larger because there could be multiple characters in a single policy narrative.

Figure 4. 3. Character Appearance over Time

Hindi Media Coverage

BEFORE GEAC APPROVAL AFTER GEAC APPROVAL AND AFTER BAN BEFORE BAN Other Other Beneficiaries 20% (3) 0% (0) Other Beneficia14% (3) 33% (1) ries 4% (1)

Beneficiaries Victims Victims Heroes 0% (0) 14% (3) Heroes 7% (1) 53% (8) 54% (12) Heroes Victims 67% (2) Villains 0% (0) Villains Villains 20% (1) 0% (0) 14% (3)

8 9

English Media Coverage

BEFORE GEAC APPROVAL AFTER GEAC APPROVAL AND Beneficiaries AFTER BAN 0% (0) Beneficiaries BEFORE BAN Other Other Victims 5% (2) 3% (1) 7% (2) 0% (0) Other 33% (1) Villains Victims 5% (2) 7% (2) Heroes Beneficiaries Villains 90% 0% (0) 10% (3) (35) Heroes Victims 67% (2) Villains Heroes 0% (0) 0% (0) 73% (22)

Character groups from previous NPF research were adapted for this study (Figure 4.4) and were classified as business/industry, government/public sector, conservationist/environment/farmer/NGO, scientist/educational, and other (Crow et al., 2017;

Shanahan et al., 2013).17 Heroes from the government/public sector were more frequent in both narratives. The Minister for Environment and Forests was extremely active in the policy process and he appears frequently in the narratives. The Minister for Agriculture and the Minister for

Science and Technology are the other two government ministers who appeared as heroes in both narratives. They were both in support of commercialization of Bt eggplant. Heroes from conservation/environment/farmer/NGO appeared more frequently in the English narratives

(n=10) than in the Hindi (n=1) narratives. Surprisingly, neither of the narratives had heroes from the industry. While the English narratives had more villains from the industry (n=3), Hindi narratives highlighted villains from the government/public sector (n=5) as well. Beneficiaries in both narratives were from conservation/environment/farmer/NGO group. Scientists appeared more frequently in Hindi (n=4) as compared to English (n=2) narratives as the ‘other’ character.

These ‘other’ characters were mostly providing information.

17 Note that previous studies (Crow et al., 2017; Shanahan et al., 2013)have a slightly different categorization. Based on a comprehensive literature review and categories that emerged during coding and analysis, the character groups have been modified so that they fit better with the policy issue under consideration.

Figure 4. 4. Heroes and Villains Across Hindi and English Media

Heroes In Hindi Media Villains In Hindi Media other other scientist/educational business/industry 0% (0) conservationist/environment/ 9% (2) 0% (0) 0% (0) farmer/NGO 0% (0) business/industry scientist/educ 17% (1) ational 23% (5)

government/p conservationi ublic sector st/environme 64% (14) nt/farmer/NG government/ O public sector 4% (1) 83% (5)

Heroes In English Media Villains In English Media business/industry other scientist/educ other 0% (0) 0% (0) ational 0% (0) 20% (1) scientist/educational 27% (16)

government/p conservationist/envir ublic sector onment/farmer/NGO business/indu conservationis 56% (33) 20% (1) stry t/environment 60% (3) /farmer/NGO 17% (10) government/ public sector 0% (0)

Since heroes appeared more frequently in both narratives, the study examined the appearance of hero groups over time (Figure 4.5). In Hindi narratives, more heroes from the government/public sector were present after GEAC approval and before ban (n=7). This aligns with the public consultations that were presided over by the Minister for Environment and

Forests who was frequently covered in the media as a hero providing solutions. In the English narratives heroes from the government/public sector were more frequent after ban (n=21) when the policy decision was discussed much more in retrospect. The second most frequent hero group was scientist/educational in both Hindi (n=5) and English narratives (n=16) and in English narratives, these were more frequent after ban (n=10) than in the other two time periods.

Conservationist/environment/farmer/NGO group appeared as hero only once in the Hindi narrative and that was after ban, while they were comparatively more frequent in the English narratives (n=10).

Figure 4. 5. Appearance of Heroes over Time Across Hindi and English Media

Heroes over Time: Heroes over Time: Hindi Media Coverage English Media Coverage 8 25 7 6 20 5 4 15 3 2 10 1 0 5 Before GEAC After GEAC After Ban Approval Approval and 0 Before Ban Before GEAC After GEAC After Ban Approval Approval and government/public sector Before Ban conservationist/environment/farmer/NGO government/public sector scientist/educational conservationist/environment/farmer/NGO other heroes scientist/educational

Given the persuasiveness of heroes in influencing opinions as depicted in previous NPF research, it is not surprising to see a high frequency of heroes across both narratives (Jones,

2014). This corresponds with the high frequency of solutions since it is heroes who typically offer solutions. There was some variation in appearance of heroes over time and which groups they represented. Heroes from the government/public sector were frequent in both, this was not the case with heroes from conservation/environment/farmer/NGO groups. This group appeared much more frequently in English than in the Hindi narratives. Since most of these groups are associated with policy actors who are more active in the urban areas, the relatively less coverage in the Hindi media might be due to its focus on issues that would appeal to the non-urban public.

Given the extreme polarization in the debate, one would expect a policy landscape with both heroes and villains pitted against each other. Hence, the starkly lower frequency of villains is surprising. Characters often advocated a solution without highlighting a villain to blame for the problem. This is an area in which further research may be helpful in terms of understanding how

characters are portrayed, whether they are instrumental in influencing policy positions, and how they are perceived by audience in non-western contexts.

Use of Evidence

As stated above, evidence is linked to the narrative element of setting. Evidence may be used for several purposes to: reflect diverging policy beliefs (McBeth et al., 2010; Shanahan et al., 2013, 2008); provide support for solutions, though this is rarely the case (Shanahan et al.,

2013, 2008); and indicate scientific uncertainty (Gupta et al., 2014; McBeth et al., 2007). NPF research has also linked evidence to other narrative elements (Radaelli et al., 2013; Schlaufer,

2016; Smith-Walter et al., 2016). Solutions need not be supported by empirical or scientific evidence since the policy issue can go beyond just the scientific aspect (McBeth, Lybecker, et al., 2014). Given all of this, one would expect evidence not being used much to support a policy preference and opposing pieces of evidence being used by coalitions to reflect diverging policy positions.

Table 4. 8. Use of Evidence over Time across Hindi and English Media

Percentage (n) Percentage (n)

Use of Evidence

Hindi (n=19) English (n=52) Before GEAC Approval: 0.00 (0) 3.85 (2) After GEAC Approval and Before Ban: 21.05 (4) 11.54 (6) After Ban: 21.05 (4) 26.92 (14) No Evidence Used 57.89 (11) 57.69 (30) Total 100.00 (19) 100.00 (52)

Use of Evidence – Findings and Discussion: Policy narratives containing evidence appeared more frequently after regulatory approval in both the Hindi (n=8) and English (n=20) media

(Table 4.8). The study also examined the types of evidence provided in support of a policy preference across Hindi and English media over time (Table 4.9). While evidence was not cited frequently across all three time periods in the Hindi media (n=11), this holds true for English media after GEAC approval (n=30). Statistical data were used in the Hindi media but do not appear in the English media, while evidence directly from scientists is more frequently cited in the English media across all three time periods (n=18) than in the Hindi media (n=3).

Table 4. 9. Types of Evidence provided over time

• No Evidence Used • No Evidence Used (5) • No Evidence Used (3) (3) • Statistical Data (2) • Evidence from scientists • Evidence from scientists (1) (2) • Descriptive Data (1) • Evidence from regulator/government (1)

Hindi Media

BEFORE GEAC AFTER GEAC APPROVAL AFTER BAN APPROVAL AND BEFORE BAN

English Media

• Evidence from • No Evidence Used (9) • No Evidence Used (21) scientists (2) • Evidence from scientists (4) • Evidence from scientists • Evidence from (12) regulator/government (1) • Evidence from • Descriptive Data (1) regulator/government (2)

Across both Hindi and English media, a higher frequency of narratives did not contain evidence in support of a policy preference. This is consistent with findings from previous NPF research indicating that evidence is rarely provided (Shanahan et al., 2013, 2008) and also that more frequent use of evidence is not necessarily connected with successful narrative strategies that influence policy outcomes (McBeth et al., 2007; Shanahan, Jones, et al., 2011). Lack of evidence is also consistent with NPF findings that policy solutions often go beyond scientific justifications or evidence (McBeth, Lybecker, et al., 2014). Given the contentiousness of the issue of agricultural biotechnology that stems not only from contradictory scientific evidence but also issues related to social, cultural, and political elements, it is not surprising that majority of the narratives do not contain scientific evidence.

Conclusion

The goal of this study was to apply the Narrative Policy Framework across different languages to examine the assumption that distinctive, generalizable narrative elements exist across multiple languages. Specifically, the analysis explored whether narrative elements are transportable outside of the English language (e.g., are villains used in non-English language narratives?) and whether narrative elements are used similarly in different languages on the same policy issue over time (e.g. a higher frequency of heroes aligns with a higher frequency of solutions) to assess the stability of policy narratives across linguistic contexts. The comparison between Hindi and English policy narratives on the same policy issue provides insight into the composition and structure of policy narratives across languages. The narrative elements examined in this study namely, policy problems, solutions, characters, and evidence were prevalent in the Hindi narratives, which lends support to the transportability of narrative elements and can be used to study issues in policy subsystems in varied linguistic contexts.

The difference in frequency of coverage across the two datasets indicates that the issue foci of the two-language media may differ. As discussed earlier, national level newspapers in the regional language tend to cover local as opposed to global stories. Since agricultural biotechnology was covered more frequently in the English media, it could indicate that the issue is more prevalent among the elite who are the majority consumers of English media, while the regional language media is more focused on the non-urban public. The difference in issue coverage could also be explained in terms of the difference in interests between the English and

Hindi readers and that India’s Hindi newspapers at the national level employ local journalists to generate news from rural areas to cater to readers interested in local news. Thus, examining

English and regional language policy narratives are important to get a more comprehensive sense

of the prevalence of particular policy issues at local, regional, and national scales. This study only included events in the English dataset that were prevalent in the Hindi dataset. Future research exploring the events that were ignored in the Hindi media may help to further understand the differences in the composition of policy narratives across the two languages.

The quantitative comparison of the narrative elements did not reveal much difference in the frequency of their appearance. Policy problems and solutions were frequently highlighted by both media. Both media contained few pieces of evidence in support of a policy preference and both had a higher frequency of heroes. The differences lay in the frequency of these narrative components over time illustrated through how these elements appeared in the three time periods under consideration. Additionally, though problems, solutions, characters, and use of evidence may not have differed much in frequency, they differed in terms of which problems were defined, what solutions were provided, who the characters were, and what kinds of evidence were used in the two languages and showed that use of narrative elements varied over time in

Hindi and English narratives and use of some narrative elements did not align with expectations from previous NPF research such as the lack of villains in a contentious policy issue, though appearance of certain narrative elements such as heroes and solutions aligned with prior NPF expectations.

This study contributes to the growing number of NPF studies that have been conducted in non-US contexts and lends support to the transportability of the framework, especially in terms of narrative elements. Given the predominance of non-English languages in the developing and less developed regions of the world, it is imperative to examine whether NPF, which was primarily developed within a North American context and within the English language, can actually be applied to policy narratives in different linguistic and national contexts. Additionally,

a singular focus on policy narratives in English may result in the marginalization of portrayals of narrative elements outside of the English language. As the literature discussed above shows, non-

English policy narratives are consumed much more by stakeholders and non-urban public in the

Indian context. Hence, it is salient to study non-English policy narratives in such linguistically diverse policy subsystems. Such an examination may also contribute to our understanding of how non-English policy narratives can differ in the way information is delivered and what information is delivered to the public. For example, this study showed differences in the kinds of problems and solutions that were discussed more frequently across the two languages, along with differences in the types of evidence used. Although further research is needed to understand the influences behind these variations such as the cultural context, ideological differences, socio- economic differences, and potential capacity constraints of Hindi and English media organizations, this study helps fill an important gap in comparative NPF research. More comparative linguistic studies in a variety of national contexts will help establish the generalizability of the Narrative Policy Framework and its responsiveness to linguistic and cultural specificity in different policy subsystems.

References

Bagla, P., & Stone, R. (2013). Scientists clash swords over future of GM food crops in India. Science, 340(May), 539–540. https://doi.org/10.1126/science.340.6132.539

Bandopadhyay, R., Sinha, P., & Chaudhary, B. (2012). Is Bt-brinjal ready for future food?--A critical study. Indian Journal of Biotechnology, 11(2), 238–240.

Baumgartner, F.R. & Jones, B.D. (2009). Agendas and Instability in American Politics. Chicago, IL: University of Chicago Press.

Boykoff, M.T. (2011). Who Speaks for the Climate: Making Sense of Media Reporting on Climate Change. Cambridge, UK: Cambridge University Press.

Crow, D. A., Berggren, J., Lawhon, L. A., Koebele, E. A., Kroepsch, A., & Huda, J. (2016). Local media coverage of wildfire disasters: An analysis of problems and solutions in policy narratives. Environment and Planning C: Government and Policy. https://doi.org/10.1177/0263774X16667302

Crow, D. A., & Lawlor, A. (2016). Media in the Policy Process: Using Framing and Narratives to Understand Policy Influences. Review of Policy Research, 33(5), 472–491. https://doi.org/10.1111/ropr.12187

Crow, D. A., Lawhon, L. A., Berggren, J., Huda, J., Koebele, E., & Kroepsch, A. (2017). A Narrative Policy Framework Analysis of Wildfire Policy Discussions in Two Colorado Communities. Politics & Policy, 45(4), 626–656. https://doi.org/10.1111/polp.12207

Fowler, R. (2013). Language in the News: Discourse and Ideology in the Press. Routledge.

Friedlander, P., Jeffrey, R., & Seth, S. (2001). “Subliminal Charge”: How Hindi-Language Newspaper Expansion Affects India. Media International Australia Incorporating Culture and Policy, 100(1), 147–165. https://doi.org/10.1177/1329878X0110000114

Ghosh, A. and Boykoff, M. (In Press). Framing sustainability and climate change: Interrogating discourses in vernacular and English-language media in Sundarbans, India

Government of India (GoI). Constitutional Provisions: Official Language Related Part-17 of The Constitution of India. Department of Official Language, Government of India. Accessed on March 22, 2017.

Gupta, K., Ripberger, J. T., & Collins, S. (2014). The Strategic Use of Policy Narratives: Jaitapur

100

and the Politics of Siting a Nuclear Power Plant in India. In M. D. Jones, E. A. Shanahan, & M. K. McBeth (Eds.), The Science of Stories: Applications of the Narrative Policy Framework in Public Policy Analysis (1st ed., pp. 89–106). New York: Palgrave Macmillan.

Haspelmath, M. (2010). Comparative concepts and descriptive categories in crosslinguistic studies. Language, 86(3), 663–687.

Herring, R. J. (2006). Why did “Operation Cremate Monsanto” fail? Science and class in India’s great terminator-technology hoax. Critical Asian Studies, 38(4), 467–493.

Herring, R. (2012). State science and its discontents: why India’s second transgenic crop did not follow the path of Bt cotton’. In Weihenstephaner Socio-Economic Seminar, Center of Life and Food Sciences Weihenstephan, Technische Universität M ünchen (Vol. 13).

Herring, R. J. (2015). State science, risk and agricultural biotechnology: Bt cotton to Bt Brinjal in India. Journal of Peasant Studies, 42(1), 159–186.

Huda, J. (2018). An Examination of Policy Narratives in Agricultural Biotechnology Policy in India. World Affairs, 181(1), 42–68.

Huda, J. (2019). Policy Narratives across Two languages: A Comparative Study using the Narrative Policy Framework. Review of Policy Research, 36(4), 523-546.

Indian Readership Survey (IRS). (2017). Retrieved from http://mruc.net/uploads/posts/a27e6e912eedeab9ef944cc3315fba15.pdf

Iyengar, S. (1990). Framing responsibility for political issues: The case of poverty. Political Behavior, 12(1), 19–40.

Jainl, B. (2014, June 21). Nearly 60% of Indians speak a language other than Hindi. Times of India. New Delhi.

Jeffery, R. (2000). India’s Newspaper Revolution. London: C. Hurst.

Jones, M. D. (2013). Cultural Characters and Climate Change: How Heroes Shape Our Perception of Climate Science. Social Science Quarterly, 95(1), 1–39. https://doi.org/10.1111/ssqu.12043

Jones, M. D. (2014). Communicating Climate Change: Are Stories Better than “Just the Facts”? Policy Studies Journal, 42(4), 644–673. https://doi.org/10.1111/psj.12072

Jones, M. D., & McBeth, M. K. (2010). A Narrative Policy Framework: Clear Enough to Be Wrong? Policy Studies Journal, 38(2), 329–353. Retrieved from http://search.proquest.com/docview/210543073?accountid=452

101

Jones, McBeth, & Shanahan, (2014). Introducing the Narrative Policy Framework. In The Science of Stories: Applications of the Narrative Policy Framework in Public Policy Analysis. Eds. Jones, M. D., Shanahan, E. A., & McBeth, M. K. New York, NY: Palgrave Macmillan.

Jones, M. D., Shanahan, E. A., & McBeth, M. K. (2014). The Science of Stories: Applications of the Narrative Policy Framework in Public Policy Analysis. New York, NY: Palgrave Macmillan.

Kingdon, J.W. (2003). Agendas, Alternatives and Public Policies. New York, NY: Longman.

Krippendorf, K. (2004). Content Analysis: An Introduction to Its Methodology. Thousand Oaks, CA: Sage Publications.

Laursen, L. (2012). Monsanto to face biopiracy charges in India. Nature Biotechnology, 30(1), 11.

Lawton, R., & Rudd, M. (2014). A Narrative Policy Approach to Environmental Conservation. AMBIO, 1–9. https://doi.org/10.1007/s13280-014-0497-8

McBeth, M. K., Jones, M. D., & Shanahan, E. A. (2014). The Narrative Policy Framework. In P. A. Sabatier & C. Weible (Eds.), Theories of the policy process (3rd ed.). Boulder, CO: Westview Press.

McBeth, M. K., Lybecker, D. L., & Husmann, M. A. (2014). The Narrative Policy Framework and the Practitioner: Communicating Recycling Policy. The Science of Stories: Applications of the Narrative Policy Framework in Public Policy Analysis, 45.

McBeth, M. K., Shanahan, E. A., Hathaway, P. L., Tigert, L. E., & Sampson, L. J. (2010). Buffalo tales: interest group policy stories in Greater Yellowstone. Policy Sciences, 43(4), 391–409.

McMorris, C., Zanocco, C., & Jones, M. (2018). Policy Narratives and Policy Outcomes: An NPF Examination of Oregon’s Ballot Measure 97. Policy Studies Journal.

Merry, M. K. (2016). Constructing Policy Narratives in 140 Characters or Less: The Case of Gun Policy Organizations. Policy Studies Journal, 44(4), 373–395. https://doi.org/10.1111/psj.12142

102

Miles, M. B., Huberman, A. M., Huberman, M. A., & Huberman, M. (1994). Qualitative data analysis: An expanded sourcebook. sage.

Neyazi, T. A. (2010). Cultural imperialism or vernacular modernity? Hindi newspapers in a globalizing India. Media, Culture & Society, 32(6), 907–924.

Neyazi, T. A. (2011). Politics after Vernacularisation: Hindi Media and Indian Democracy. Economic and Political Weekly, 46(10), 75–82. Retrieved from http://www.jstor.org/stable/41151945

Ninan, S. (2007). Headlines from the Heartland: reinventing the Hindi public sphere. New Delhi: Sage.

Pierce, J. J., Smith-Walter, A., & Peterson, H. L. (2014). Research Design and the Narrative Policy Framework. In M. D. Jones, E. A. Shanahan, & M. K. McBeth (Eds.), The Science of Stories: Applications of the Narrative Policy Framework in Public Policy Analysis (pp. 27–44). New York: Palgrave Macmillan.

Radaelli, C. m., Dunlop, C. a., & Fritsch, O. (2013). Narrating Impact Assessment in the European Union. European Political Science, 12(4), 500–521. https://doi.org/10.1057/eps.2013.26

Ramaswami, B., & Pray, C. E. (2007). India: confronting the challenge–the potential of genetically modified crops for the poor. The Gene Revolution: GM Crops and Unequal Development, 156–174.

Rao, S. (2009). Glocalization Of Indian Journalism. Journalism Studies, 10(4), 474–488. https://doi.org/10.1080/14616700802618563

Sabatier, O. A., Hunter, S. & Mclaughlin, S. (1987). The Devil Shift: Perceptions and Misperceptions of Opponents. Western Political Quarterly 40 (3): 449-476.

Sahay, U. (2006). Making news: Handbook of the media in contemporary India. Oxford University Press.

Samuels, J. (2013). Transgene flow from Bt brinjal: a real Risk? Trends in Biotechnology, 31(6), 332–334.

Scheufele, D.A. (2000) Agenda-setting, priming, and framing revisited: Another look at cognitive effects of political communication. Mass Communication and Society 3: 297– 316.

Scheufele, D.A. & Tewksbury, D. (2007). Framing, agenda setting, and priming: The evolution of three media effects models. Journal of Communication 57: 9–20.

Schlaufer, C. (2016). The Narrative Uses of Evidence. Policy Studies Journal, n/a-n/a.

103

https://doi.org/10.1111/psj.12174

Secretariat of the Convention on Biological Diversity. (1992). Retrieved from https://www.cbd.int/doc/legal/cbd-en.pdf

Shanahan, E. A., Jones, M. D., & McBeth, M. K. (2011). Policy Narratives and Policy Processes. Policy Studies Journal, 39(3), 535–561. Retrieved from http://search.proquest.com/docview/887282322?accountid=452

Shanahan, E. A., McBeth, M. K., & Hathaway, P. L. (2011). Narrative Policy Framework: The Influence of Media Policy Narratives on Public Opinion. Politics & Policy, 39(3), 373– 400. https://doi.org/10.1111/j.1747-1346.2011.00295.x

Shanahan, E. A., McBeth, M. K., Hathaway, P. L., & Arnell, R. J. (2008). Conduit or contributor? The role of media in policy change theory. Policy Sciences, 41(2), 115–138.

Specter, M. (2014). Seeds of Doubt. The New Yorker. Retrieved from http://www.newyorker.com/magazine/2014/08/25/seeds-of-doubt

Stone, D. A. (2011). Policy paradox: The art of political decision making. London and New York: WW Norton New York.

Tosun, J., & Schaub, S. (2017). Mobilization in the European Public Sphere: The struggle over genetically modified organisms. Review of Policy Research, 34(3), 310–330.

United States Department of Agriculture (USDA). Retrieved on August 25, 2017 from https://www.usda.gov/topics/biotechnology/biotechnology-frequently-asked-questions- faqs

Wenzelburger, G., & König, P. D. (2017). Different by Design? Analyzing How Governments

104

Justify GMO Liberalization through the Lens of Strategic Communication. Review of Policy Research, n/a-n/a. https://doi.org/10.1111/ropr.12237

105

CHAPTER 5: SOURCES OF EVIDENCE FOR RISKS AND BENEFITS IN AGRICULTURAL BIOTECHNOLOGY POLICY IN INDIA: EXAMINING SETTING AND PLOT IN POLICY NARRATIVES

A December 2018 article co-authored by the father of India’s Green Revolution, MS

Swaminathan18 raised concerns about the environmental harm from genetically modified (GM) crops and claimed that some GM crops can damage genetic information in humans. The scientific community saw it as a thinly-veiled attack against agricultural biotechnology from a well-known proponent19 (Mehta, 2018; Vembu, 2018). This example highlights that agricultural biotechnology continues to remain contentious in India despite India being the fifth largest GM crop producing country in the world (ISAAA, 2017).

In controversial areas such as introduction of new technologies, generating socially relevant knowledge about evidence and risks becomes important for mediating political and normative conflicts. Fundamental questions about sources of expertise and problem framing become important (Gupta, 2011). To fully understand the story told by proponents and opponents of agricultural biotechnology in the policy process, scholars must investigate the sources of expertise that they rely on and how risks are articulated by opposing coalitions.

Although risk and evidence in genetically modified organisms (GMOs) have been studied extensively (Herring, 2008, 2015; Ho, 2000; Pinstrup-Andersen & Schiøler, 2003; Schurman &

Munro, 2006), research indicates that using the Narrative Policy Framework (NPF) to examine the public discourse on GMOs can be a rewarding enterprise (Tosun & Schaub, 2017). After all,

“there is no such thing as “just the facts” since these are embedded in narratives that shape our

18 MS Swaminathan led the Green Revolution in India in the 1960s and ’70s that helped increase agricultural productivity in the country dramatically, saving it from famine (Mehta, 2018). 19 Following criticism that the article was scientifically flawed, MS Swaminathan designated his co-author, PC Kesavan, a geneticist and radiation scientist as the sole author thereby distancing himself from the controversy.

106

cognitive processes and contribute to public policy decisions (Schlaufer, 2016, p 108). In NPF scholarship, scientific and technical information is not merely interpreted as numbers and data but incorporated into stories (Weible & Schlager, 2014). Recent NPF research has increasingly focused on the role of evidence (Mosley & Gibson, 2017; Schlaufer, 2016; Shanahan, Jones,

McBeth, et al., 2018; Smith-Walter et al., 2016) and risk framing (Lawlor & Crow, 2018; Raile et al., 2018).

Policy scholars have contended that individual cognitive processes are important given the limited ability of policymakers to tackle massive amounts of data and rhetoric (Jones, 2003) and personal values necessarily influence policy decisions (Mosley & Gibson, 2017; D. A.

Stone, 1997). While “policymaker uncertainty may be best addressed with scientific evidence, policymaker ambiguity may be best addressed with narrative persuasion” (Mosley & Gibson,

2017, p 700). Thus, understanding how evidence and narrative shape one another in the policy process is important. NPF focuses on the central role of narratives in the policy process. Using the NPF, this paper examines how proponents and opponents of agricultural biotechnology use evidence to support their claims about risks and benefits. This may help unravel policy process challenges for governments assessing risks and benefits in the face of contradictory evidence.

This paper examines the policy process for a genetically modified variety of eggplant in

India called Bt eggplant, which was widely perceived as potentially India’s first GM food crop, but a decade of field trials led to an indefinite moratorium. In the Bt eggplant discourse,

“uncertainty was constructed as ‘risk,’ … which … trumped official scientific evidence”

(Herring, 2014, p 207) making it an appropriate case study to examine evidence and risk. The paper begins with a theoretical discussion on the use of evidence and risk within NPF and public policy scholarship. The case study is then discussed including how evidence for risks and

107

benefits in agricultural biotechnology has been studied with a focus on the Bt eggplant case.

Methods are then discussed including research design and data collection followed by findings and discussion. The analysis reveals that stakeholders use different sources of evidence and proponents de-emphasize risks and exclusively highlight benefits while opponents invoke multi- dimensional risk. Lastly, risk perceptions of stakeholders are influenced by moral notions of risk.

NPF and Public Policy Scholarship: Sources of Evidence and Perception of Risks and Benefits

Given the importance of narratives in the policy process, the Narrative Policy Framework systematically analyzes the role and influence of narratives on policy outcomes and its strategic use by opposing coalitions. NPF has four narrative core elements: setting (policy problems situated in a specific policy context), characters (heroes, villains, victims, beneficiaries, among others), plot (arc of action), and moral of the story (policy solution) (Shanahan et al., 201820).

Setting and plot in policy narratives discuss physical, temporal, or governmental aspects of policy problems and causal relationships underlying the problem (Lawlor & Crow, 2018).

Setting is the context of the policy problem consisting of “low-contestation ‘facts’” generally agreed upon by policy actors, different types of evidence and “indicators, legal and geographic characteristics and any other policy consequential element of the environment in which the policy exists” (Jones & Radaelli, 2015, p 341). It includes ‘known’ facts including science or evidence about the problem (Crow & Jones, 2018). Examining setting in the context of risks and hazards, Lawlor & Crow (2018) discuss how setting “may emerge as a discussion of evidence about the known or unknown risks that a community or a nation faces from” hazards. Risk information may include objective measurement of risk or may contrast “emotional or

20 This citation reflects the latest NPF theory but is built on previous NPF works. See Jones & McBeth, 2010; McBeth et al., 2014; Shanahan, Jones, & McBeth, 2011; Shanahan et al., 2013.

108

empathetic appeals to the audience” about risk (p 848-9). Thus, setting in NPF includes evidence that helps establish the risk under contention or consideration, thereby making both evidence as well as risks crucial to understanding the setting in a policy narrative in policy subsystems characterized by risk. Furthermore, the plot in policy narratives also includes evidence pertaining to human ability to address the policy problem such as existing policies or those that could be implemented to mitigate risk in agricultural biotechnology and bring about policy change.

Considerable NPF research has focused on refining narrative elements related to characters, problem definitions, and moral of the story (Jones, 2014; Shanahan, Jones, McBeth,

& Lane, 2013; Shanahan et al., 2018). Recent NPF research has examined evidence (Mosley &

Gibson, 2017; Schlaufer, 2016; Smith-Walter et al., 2016) and risk (Lawlor & Crow, 2018) but not always explicitly in relation to plot and setting. In policy areas characterized by risk and where evidence is contested such as the introduction of novel technologies, risk and evidence are both essential components of the plot and setting of policy narratives. These may help understand the cause of the policy problem, who or what is to blame for the problem (villains), who may provide solutions to the problem (heroes), and the way in which evidence is employed in order to narrate the risk. Additionally, evidence can be constructed in a number of ways and includes not just risks but also benefits (societal, economic, and other types), which are an important aspect of risk communication and a policy consequential element as certain governments mandate the evidence of benefits to approve certain risk technologies. Though evidence and risk have been studied in NPF research, benefits have not been examined but are essential components of the plot and setting since benefits may support the moral of the story and help understand who benefits (beneficiaries) from adoption of risk technologies.

109

Evidence in Public Policy Scholarship and NPF: Situating evidence within the discipline of public policy, Smith-Walter et al. (2016) show that evidence has largely been discussed within utilization literature in particular and policy process literature in general. Based on Young,

Ashby, Boaz, & Grayson (2002), they discuss five basic models of utilization in the policy process: (i) the knowledge-driven model – evidence drives policy directly (ii) the problem- solving model – policy drives research (iii) the interactive model – research and policy are interactive and “shaped within policy communities of diverse actors” (iv) the political/tactical model – “actors strategically use evidence to win” and (v) the enlightenment model – research stands apart from policy and affects it indirectly (pp 1057-8).

The relationship between evidence and policy is more holistic in the policy process literature since evidence is situated among other important factors in the policy process. Based on Weible's (2008) examination of expert-based information, Smith-Walter et al. (2016) discuss evidence and public policy in context of four major policy process theories. In multiple streams theory (Kingdon, 2003), problems and solutions are identified using science; for the science to be effective, there needs to be a skillful policy entrepreneur; and “entrepreneurs use science for political gain” (p 1058). In punctuated equilibrium theory (Baumgartner & Jones, 1993), there are four implications linked with evidence: the causal driver is the “pace with which actors process expert-based information”; disproportionate information processes create, maintain, destroy, or alter policy images; “science affects the expansion of conflict and the mobilization of resources”; and “science contributes to all kinds of policy change” (p 1058). In social construction and policy design theory, science may be understood to the extent that the science community and policy community are respectively unified and “expert-based information may provide risks and opportunities to powerful groups” (p 1059). Lastly, in the Advocacy Coalition

110

Framework, coalitions use expert-based information to recruit allies and fight opponents, and policy-oriented learning based on evidence likely occurs within one coalition (p 1059).

Given the “political challenges between evidence-based decision making (Pew-

MacArthur Results First Initiative 2014) and science denial (Rosenau 2012),” NPF studies increasingly include an examination of evidence in policy narratives where evidence is treated more like a prop that “may become contested or the focal point of the policy narrative”

(Shanahan et al., 2018, p 176). NPF research has used the terms ‘science’ and ‘evidence’ to designate data used in support of policy positions. However, since the 1990s “evidence-based” policymaking has become a more prominent term in the policy process literature and more recent

NPF research has used the term ‘evidence’ (McBeth, Jones, et al., 2014b; Mosley & Gibson,

2017; Schlaufer, 2016; Shanahan et al., 2013; Smith-Walter et al., 2016).

In reviewing the major policy process theories, Smith-Walter et al. (2016) find that evidence is predominantly discussed in terms of scientific information in policy process literature. They contend that since NPF aims to understand structure and content of policy narratives and their impact on decision making, a broader understanding of evidence is needed, not just limited to scientific information. NPF research has shown that evidence reflects diverging policy beliefs (McBeth et al., 2010; Shanahan et al., 2013, 2008); is used to support the moral of the story, though rarely provided (Shanahan et al., 2013, 2008); and is used to support different narrative strategies to indicate scientific uncertainty and disagreement (Gupta,

Ripberger, & Collins, 2014; McBeth, Shanahan, Arnell, & Hathaway, 2007). Schlaufer (2016) connected evidence to all narrative elements while Smith-Walter et al. (2016) refined the use of evidence through the use of distinct categories (scientific studies, statistics, polls, ipso dictum, and legal) and Mosley & Gibson (2017) examined how different types of evidence are used at

111

specific points in the policy process to improve policy outcomes. Smith-Walter et al. (2016) emphasize the need for NPF scholars “to better understand what counts as evidence among advocacy coalitions and their specific actors within subsystems” (pp 1074-75) and to examine whether different types of evidence are used by different policy actors and whether some forms of evidence may be used more frequently in certain subsystems than others.

Risks and Benefits in Public Policy Scholarship and NPF: Risk perception appeared on political agendas in the 1960s in context of the public opposition to new technologies, especially nuclear technology (Kellens, Terpstra, & De Maeyer, 2013). Research showed a “systematic relation between the acceptance of technological risks and the perception of costs and benefits from these technologies” (Kellens et al., 2013, p 25; Starr, 1969). Social acceptance of risk was associated with perception of benefits. Risk communication has been studied alongside risk perception since communication may influence perception. Covello, Slovic, & Von Winterfeldt (1986) define risk communication as “any purposeful exchange of information about health or environmental risks between interested parties” (qtd. in Kellens et al., 2013, p 25). Since every policy context varies, it is important to adjust risk communication based on the needs of the people so that they can judge their own situation to make informed decisions.

Previous research on risk communication acknowledges that since scientists and citizens do not share a common language for risk, simply diffusing scientific information into the public domain rarely influences public opinion in varied policy areas. Scientists use the language of

“probability, uncertainty, frequency, and magnitude, whereas citizens use local language and descriptors and referents based on cultural values, cognitive biases, local knowledge, and experiences” (Raile et al., 2018, p 2). Irrespective of the language used, risk communication is

112

essential to political and policy dialogue in diverse policy domains, especially those that are characterized by risk (Golding, Krimsky, & Plough, 1992; A. Gupta, 2011; Herring, 2015;

Kellens et al., 2013; Lawlor & Crow, 2018; Wachinger, Renn, Begg, & Kuhlicke, 2013).

When risks are accurately and efficiently communicated, problem definitions and accompanying solutions may be closely connected with expert risk assessments. Efficient and accurate risk communication may also mobilize actors to influence policy change (Lawlor &

Crow, 2018). Furthermore, societal, economic, and environmental benefits are important when communicating risk technologies and influential in policymaking since governments mandate evidence of benefits to approve certain risk technologies.

Risk has only very recently been studied within the NPF and only within the hazards and disaster policy domain. Assessing how individuals respond to science-based risk messages and narrative-based risk messages with the same science language, Raile et al. (2018) put forth a narrative-based risk communication framework wherein they examined the mechanisms involved in narrative persuasion for risk communication. Linking NPF concepts of setting, plot, and moral of the story to risk framing, Lawlor & Crow (2018) investigated narrative construction of chronic versus urgent risk and its influence on policy discussions. While risk has started to garner attention from NPF scholars, benefits of potential policy solutions are still not studied, but are presumably important in risk-based decision-making and therefore an important area to develop.

Evidence for Risks and Benefits in Agricultural Biotechnology

113

United States continues to lead in the adoption of GM crops with ten crops planted in 2017, while Canada planted six. EU continues to lead the Western Hemisphere in rejection of agricultural biotechnology with a few exceptions in Spain and Portugal. Commercialization of

GM crops in Africa (thirteen countries) and Latin America (eleven countries) has increased.

Leading biotech countries in Asia and the Pacific region include India, followed by Pakistan,

China, Australia, the Philippines, Myanmar, Vietnam, and Bangladesh contributing to 10% of the global biotech crops. Lack of government approval has blocked several GM crops such as

Golden Rice, Bt eggplant, and drought tolerant and insect resistant maize in different regions. Bt eggplant, commercialized in Bangladesh is banned in neighboring India. Regulatory approval for drought tolerant and insect resistant maize, a public-private collaboration of four countries in sub-Saharan Africa remains stalled, while Brazil approved a virus-resistant bean in 2011 but the technology has not reached farmers (ISAAA, 2017).

European Union have switched positions over time and other countries like Russia have rejected the technology altogether. The committees for regulatory approval in some countries are headed by environment ministers while in others are chaired by agriculture ministers, which could also lead to divergent outcomes (Herring & Paarlberg, 2016; Paarlberg, 2001). One common thread is that evidence is invoked by both proponents and opponents (Herring, 2008, 2015; Ho, 2000;

Pinstrup-Andersen & Schiøler, 2003; Schurman & Munro, 2006). The role of evidence in the agricultural biotechnology regulatory process varies across crops and nations. Jasanoff (2005) examined variable styles of biotechnology governance in the industrialized world with a longer

114

experience with this technology. Based on research in Kenya, China, India, and Brazil, Paarlberg

(2001) put forth a multi-dimensional typology for biotechnology governance that ranges from preventive to promotional, reflecting the permissiveness of the policy regime. Across Japan,

France, and the United States, Sato (2007) noticed variation in the cultural politics of contesting evidence on biotech food safety (Herring, 2015). One lesson is that there is no such thing as ‘just the facts’ when it comes to evidence. Evidence is “filtered, weighted and selected through political processes” and variation along these lines can help explain patterns of acceptance or rejection of GM crops (p 160).

Communication of risk technologies and, specifically, of agricultural biotechnology has been studied in terms of the cultural reception, social construction, and legitimation or delegitimation in public or media discourse (Andrée, 2002; Brooks, 2005; Crawley, 2007;

Herrick, 2005; Levidow, Carr, Wield, & Von Schomberg, 1997; Levidow & Marris, 2001;

McInerney, Bird, & Nucci, 2004; Sato, 2013; Wenzelburger & König, 2017). These studies help understand controversies surrounding agricultural biotechnology with regard to their social perceptions and acceptance, role of evidence and normative standards and values in risk perception. Studies about communication of risk technologies center on risk and uncertainty and its social assessment and acceptability. Due to the complexity of the technology, perceptions of risk are malleable and what counts as acceptable risk for some stakeholders may not be the same for others (Levidow et al., 1997; Wenzelburger & König, 2017). Framing experiments show that even the terminology for biotechnology can affect its acceptance (Cacciatore, Scheufele, &

Shaw, 2012). Some studies even claim that “a risk narrative of threatened common interests

(e.g., safety, environment), based on a discourse of corporate dominance and exploitation” has contributed to the success of the anti-GMO movement (Herring & Paarlberg, 2016, p 409).

115

Because new technologies bring with them uncertainty about their risks and promises, new policy process challenges are created for governments during policy deliberation with regard to assessment of risks and benefits in the face of contradictory evidence.

India is an excellent country in which to study the GM discourse. After the green revolution in the 1960s, India became largely self-sufficient in agricultural production (Das &

Bhardwaj, 2015). However, rapid growth in population, reduction in arable land due to industrialization, and reliance on monsoons are some of the reasons leading to reduced agricultural production. Agricultural biotechnology has increasingly been publicized as the best means of bringing about a second green revolution in India (Restall, 2014). However, the matter remains contentious.

Agricultural Biotechnology in India: Evidence of Risks and Benefits in Bt Eggplant

On 26th March 2002, the regulatory body for agricultural biotechnology in India, Genetic

Engineering Approval Committee21 (GEAC), recommended the commercialization of a GM variety of cotton called Bt cotton ushering the GM debate into the Indian public sphere (Kalle &

Ejnavarzala, 2016). India is one of the top five countries with area under GM crop cultivation with Bt cotton as the only GM crop approved for commercial cultivation (ISAAA, 2017).

Despite the widespread cultivation of Bt cotton in India, other crops have not received approval for commercial cultivation.

After the widespread adoption of Bt cotton, Bt eggplant was being perceived as potentially the first food crop to be considered seriously by the regulatory system (Ramaswami

& Pray, 2007). In India, there was intense public debate nationwide regarding its commercialization. This GM crop created through the insertion of a gene cry1Ac from the soil

21 It was renamed as the Genetic Engineering Appraisal Committee in 2010.

116

bacterium (Bacillus thuringiensis) into eggplant is said to provide the plant with resistance against lepidopteran insects like the Fruit and Shoot Borer (Leucinodes orbonalis) and Fruit

Borer (Helicoverpa armigera). Eggplant is a popular vegetable crop in India among small scale farmers as well as among low income consumers, making it known as the “poor man’s vegetable crop.” It is low in calories, has a high nutritional value and water content, serves as a good source of fiber, calcium, phosphorus, folate, vitamins B and C, and is used in traditional medicine, whereas the plant itself is used as fuel (Bandopadhyay, Sinha, & Chaudhary, 2012, p

2015). However, in 2010, it could no longer be planted legally after Minister for Environment and Forests at the time rejected GEAC’s decision and placed an indefinite moratorium pending further evaluations (Bandopadhyay et al., 2012; Herring, 2012, 2015). The moratorium overturned the regulatory body’s decision to allow the commercialization of Bt eggplant.

The moratorium on Bt eggplant is often cited by GMO-opponents as a victory in support of their continued opposition to any GM crop including GM mustard which has followed Bt eggplant in the policy pipeline. The controversy surrounding Bt eggplant was intense and the dialogue often focused on risks from planting the crop. Even though Bt eggplant carried the same transgene for insect resistance as Bt cotton (cry1Ac) and underwent the same regulatory process, its risk assessment for regulation was controversial (Gupta, 2011). With Bt cotton, economic interests dominated (see Herring, 2015 for detailed discussion on Bt cotton controversy) but with

Bt eggplant the politics of risk dominated, especially given the susceptibility of food crops to

“anxiety framings” (Herring & Paarlberg, 2016, p 410).

117

The Bt eggplant policy process in India provided a unique opportunity for participatory decision-making. Following a public outcry in the wake of the GEAC approval of Bt eggplant in

2009, the Minister for Environment held a series of stakeholder consultations nationwide that were attended by farmers, civil society organizations, consumers, scientific experts, industry, environmental groups, and government officials, among others. The Minister gathered evidence from multiple sources, including views of state and regional governments and independent scientists from India and abroad. On the basis of his review of evidence from multiple sources, he imposed a moratorium on the commercialization of Bt eggplant “till such time independent scientific studies establish, to the satisfaction of both the public and professionals, the safety of the product from the point of view of its long-term impact on human health and environment”

(Ministerial Note, MoEF 2010:17).” (qtd. in Kalle & Ejnavarzala, 2016, p 32). His reasons including biodiversity risks, intellectual property rights, and vulnerability of the Indian food system were clearly articulated and displayed transparently (Herring, 2015; Ramesh, 2015).

Although risk and uncertainty have been central in risk communication, advocacy for benefits has also been relevant. Regulatory procedures often require assessment of risks as well as benefits of agricultural biotechnology. Examining GMO liberalization in Norway and the

United Kingdom, Wenzelburger & König (2017) demonstrate how governments highlighted societal and economic benefits in their communication about the policy decisions. The agricultural biotechnology policy process in India documents not just risks but also benefits.

GEAC was tasked with assessing farmer benefits through field trials at various scales with the mandate that no crop would receive approval without evidence of benefits such as economic gain to farmers (Gupta, 2011; Herring, 2015).

118

Lastly, research shows that notions of morality dominate risk perception since “people construe risk on the basis of belief systems” ( qtd. in Chong, 2005, p 618; Sjöberg, 2000). Moral notions of risk are connected to the idea of natural vs. unnatural risk. In particular, the idea of

‘unnatural risk’ is important since people hold a deep skepticism toward the unnatural. Krimsky

& Wrubel (1996) show “there is a powerful association between the concepts of “natural” and

“safe” (and inversely, between “unnatural” and “risky”) in many people’s minds” (qtd. in Chong,

2005, p 618). GMO opponents have highlighted moral concerns to reject GM crops (Shiva,

2000), while proponents have emphasized the moral imperative to make the technology available to those that want it (Chong, 2005; Nuffield Council on Bioethics, 1999). Based on a survey of eggplant farmers in India, Chong (2005) noted “the virtual absence of moral concerns among the eggplant farmers seems to limit the universality of Sjoberg’s theoretical perspective that notions of ‘‘unnatural risk’’ and ‘‘tampering with nature’’ are central to risk perception” (p 630).

Interestingly, the study indicated that risk perception among Indian farmers is mainly influenced by economic benefits especially given the uncertainty of farmer’s livelihoods in rain-fed farming systems in developing countries like India. Additionally, risks and benefits in economic terms can be directly perceived and experienced by farmers as opposed to less tangible moral or environmental risks (Chong, 2005; Wangiker, 2004). Given the contentiousness of the evidence for risks and benefits and the prevalence (or lack thereof) of moral notions of risk (in the context of natural vs. unnatural risk), this study examines these variables using the NPF.

Methods

Case study research requires one to gather multiple sources of data, to triangulate when possible, and to be transparent and clear (Miles & Huberman, 1994; Yin, 1994). This section outlines data sources for the research. The next section triangulates the findings based on the data

119

and provides example quotations and interview questions to make the methodology transparent and clear.

Data Collection Methods: Data were collected from the following sources for the case study research presented here and analyzed using an NPF-focused codebook.

Documents: First, all policy process documents were gathered through online searches to provide contextual policy information beginning from mid-2000 when Bt eggplant was being perceived as potentially the first GM food crop to be considered seriously by the regulatory system

(Ramaswami & Pray, 2007). Second, media data were collected from the leading Hindi and

English newspapers in India and were analyzed using an NPF focused codebook using NPF coding procedures22. To capture breadth of coverage, sampling timeframe was three years before and after the ban was imposed on Bt eggplant: February 9, 2007 - February 9, 2013. This allows us to examine the kind of evidence and risk/benefit information that was used while the policy was being actively discussed in the media.

Interviews: Third, semi-structured interviews were conducted in June-August 2017 as per guidelines provided by Rubin & Rubin (2005). 41 stakeholders were contacted and 23 were

22 For the English media dataset: A total of 1,212 articles (397 from Times of India and 815 from Hindustan Times) were downloaded from the ProQuest News and Newspaper database using search terms: “Bt brinjal” OR “Bt eggplant” OR “genetically modified” OR “agricultural biotechnology.” Duplicate articles or those not focusing on Bt eggplant/brinjal or not written in narrative form were removed from the dataset leading to a final English media dataset of 227 articles (87 from Times of India and 140 from Hindustan Times) that were coded initially. Based on the ‘policy narrative’ definition (Shanahan, Jones, McBeth, & Radaelli, 2018), media articles without a policy referent and a character were removed leading to 171 articles (Times of India - 51 and Hindustan Times - 120). For the Hindi media dataset, archives of the New Delhi edition of Dainik Jagran were accessed in person at the Noida office for the same timeframe as the English media coverage. After a manual search of the dataset, articles focusing on Bt eggplant (बीट$ बग& न) were collected (n=20) and only policy narratives were included in the final dataset (n=19).

120

interviewed. After a preliminary review of each transcript in its entirety, only those interview transcripts that adhered to the policy narrative definition (a character and policy referent) were included in this analysis (n=20). Interviews were conducted with stakeholders in agricultural biotechnology who are currently or were formerly affiliated with non-profit organizations (n=8), industry (n=4), research organizations and public universities (n=5), and government including regulatory bodies (n=3). Stakeholders were identified through an analysis of media coverage, so their public position on the issue was known prior to the interview. The interview attempted to elicit their policy narratives that they had used during the policy debate with a focus on specific narrative elements. The interview participants have largely remained active in the policy subsystem with many currently involved in the GM mustard policy process. Thus, it is the researcher’s hope that they recalled the Bt eggplant policy process as accurately as humanly possible. To maximize consistency of coding and analysis, interview transcripts were coded using a constant comparative approach in NVivo software (Miles & Huberman, 1994). In addition to NPF-related questions on policy problems, solutions, characters, etc., interview participants were asked specific questions related to:

1) Use of evidence: What kinds of information about agricultural biotechnology/Bt eggplant

in general do you pay attention to? Where do you get that information from?

2) Risk/benefit information: How do you assess risks or benefits associated with agricultural

biotechnology in general and Bt brinjal in particular?

After preliminary analysis, further categories for sources of evidence and risk and benefit information were developed and subsequently coded for in-depth analysis. These will be discussed in more detail in the findings section below.

121

Policy documents were used to develop the timeline and the case description. Media data helped to establish the dominant policy narratives and also to identify interview subjects.

Interview data were then used to delve into the nuances of the policy narratives.

NPF has primarily drawn on quantitative methods such as surveys, content analysis, and experiments to fulfill its aspiration to be ‘clear enough to be wrong.’ But Jones & Radaelli

(2015) argue that NPF’s theoretical structure does not inherently discriminate between quantitative and qualitative methods (Gray & Jones, 2016). Integrating qualitative methods into

NPF’s theoretical apparatus could prove useful especially when it comes to discussion of normative policy concerns such as presence of risks and benefits. The utility of qualitative methods to the NPF is demonstrated in recent research (Gray & Jones, 2016; McMorris et al.,

2018; Mosley & Gibson, 2017). Qualitative research contributes in numerous ways, particularly to “build theory, validate constructs, and provide novel insight” (Mosley & Gibson, 2017, p

703). This study attempts to provide novel insight to explain elements in the policy process such as what kind of evidence is used by policy actors to advocate for their policy preference and to support their claims about risk and benefits about agricultural biotechnology. By focusing on the policy process for Bt eggplant in India, this study capitalizes on the strengths of the case study method used here by relying on multiple sources of data to triangulate findings and to be transparent and clear (Yin, 1994).

The media and interview data were first divided into two categories: pro-Bt eggplant (in support of commercialization) and anti-Bt eggplant (opposed to commercialization) policy narratives23. An excerpt from a pro-Bt eggplant media article is below:

In addition to the advantages of reduction in pesticide usage, minimization of yield loss and economic benefits, experts believe that Bt brinjal has a significant role in improving the nutrition

23 Policy narratives contain a minimum of one character and a policy referent (Shanahan, Jones, McBeth, et al., 2018)

122

status of the masses, since it is one of the largest produced and consumed vegetables in the country. (Nagarsekar, 2009)

An excerpt from an anti-Bt eggplant media article is below:

Pushpa M Bhargava, appointed to the GEAC by the apex court after a case in which the clearance process was challenged, has warned that enormous scientific literature was ignored in a haste to clear the first genetically modified food crop in India. (Sethi, 2009)

Policy narratives in the media focusing on Bt eggplant did not always emphasize a distinct policy preference in support or against Bt eggplant. These are classified as incomplete narratives and excluded from the current dataset24. Only pro- and anti-Bt eggplant narratives are analyzed

(n=99).

Some interview participants did not always hold a specific position on a particular GM crop but were against or in support of agricultural biotechnology in general. For instance, some interview participants opposed the commercialization of Bt eggplant but were not against agricultural biotechnology per se (classified as Anti-Bt eggplant, n=9).

I think a certain kind of agricultural biotechnology has a great role to play in establishing sustainable farming. … with Bt brinjal … it’s certainly a risky technology. It’s certainly a technology that’s not in the control of farmers. … takes away choices for both farmers and consumers. … for government. (Anti-08)

I didn’t say anything against biotech. My limited concern was Bt Brinjal … it creates some issues, we have to address that. … don’t close your doors to science, don’t close your doors to field trials but be cautious when it comes to commercialization particularly when it comes to food crops and particularly where India is the center for genetic diversity. (Anti-02)

In contrast, those who supported Bt eggplant commercialization, also supported agricultural biotechnology in general. They did not always differentiate between specific GM crops (Pro-Bt eggplant, n=11).

I have personal belief that the technology is safe. … Government should commercialize Bt brinjal straight away (Pro-01)

24 see Huda (2018) for detailed information on incomplete narratives in NPF

123

We need to understand that biotechnology has something to offer to everyone. … Bt is something that is bacillus thurigensis which is soil borne bacteria … it has [a] history of safe use (Pro-03)

I feel that there's a desperate need for increasing productivity for which you need agricultural biotechnology. … the gene used from the cry1ac... whatever genes were used have been used extensively. … So, I don't think there can be [a] real issue. I don't see any reason why we should not go for [it]... (Pro-09)

Based upon the above categorization of media data and interview subjects into pro- and anti-groups, this study examined the setting and plot in the commercialization process of Bt eggplant and is guided by the following research question:

RQ: How do proponents and opponents of agricultural biotechnology use evidence to

support their claims about risks and benefits?

To examine the research question, this paper first provides expectations from proponents (Pro-) and opponents (Anti-) regarding the use of evidence, sources of evidence, perception of risks vs. benefits, and moral notions of risks that will guide the empirical analysis. These expectations

(outlined in Table 5.1) have been developed from an in-depth analysis of the case study literature as well as a preliminary review of the media and interview data and interactions with experts in the field of agricultural biotechnology.

124

Table 5. 1. Expectations

Narrative Element Expectations Pro-Bt Eggplant Anti-Bt Eggplant Use of Evidence Proponents and opponents provide evidence to support their position (Herring, 2015) Sources of Evidence Proponents privilege peer - Opponents highlight other reviewed scientific evidence sources of evidence in addition over other sources to peer -reviewed scientific evidence Perception of Risks vs. Benefits 1. Proponents de-emphasize the 1. Opponents invoke risks multidimensional risk including 2. Proponents highlight benefits environmental, human health, outweigh the risks ethical, and religious considerations (Herring & Paarlberg, 2016; Tosun & Schaub, 2017) 2. Opponents claim there are no benefits from agricultural biotechnology Moral Notions of Risk Proponents consider genetic Opponents invoke notions of engineering as natural or part of ‘unnatural risk’ and ‘tampering agriculture with nature’ (Sjöberg, 2000)

The following section examines whether these expectations align with the data.

Findings and Discussion

Use of Evidence in Agricultural Biotechnology

Findings in the Media Data:

An analysis of evidence in the media coverage on Bt eggplant (Table 5.2), three years before and after the policy decision to impose a moratorium was taken, shows that policy narratives in support of Bt eggplant tended to use more evidence (61.54 percent, n=16) than those against Bt eggplant (43.84 percent, n=32). When evidence was cited in the media narratives, both pro- (62.5 percent, n=10) and anti- (59.38 percent, n=19) Bt eggplant narratives had a higher frequency of evidence provided directly from scientists. These included direct quotes from scientists as well as summaries and paraphrases. Pro-Bt eggplant narratives also included more evidence directly from the regulator or the government (18.75 percent, n=3) than

125

anti-Bt eggplant narratives (3.13 percent, n=1), which provided descriptive data (not attributed to scientists/regulator/government) much more frequently (37.51 percent, n=12).

Table 5. 2. Evidence in Media Data (n=99)

Pro-Bt Eggplant (n=26) Anti-Bt Eggplant (n=73) Use Evidence 61.54 (16) 43.84 (32) No Evidence 38.46 (10) 56.16 (41) Total 100.00 (26) 100.00 (73) Type/Source of evidence Pro (n=16) Anti (n=32) Evidence provided directly from 62.5 (10) 59.38 (19) scientist(s) – quotes/summary/paraphrase Evidence provided directly from 18.75 (3) 3.13 (1) regulator/government – quotes/summary/paraphrase Statistical Data (but no direct evidence 6.25 (1) 0.00 (0) from scientists/regulator/government) Descriptive Data (but no direct evidence 6.25 (1) 34.38 (11) from scientists/regulator/government) Descriptive and Statistical Data (but no 6.25 (1) 3.13 (1) direct evidence from scientists/regulator/government) Total 100.00 (16) 100.00 (32)

Findings in the Interview Data:

In order to obtain information on evidence from interview participants, they were asked about what kind of information about agricultural biotechnology in general and Bt eggplant in particular they paid attention to and where they got this information from. Such a broad question allowed participants more room to discuss their views on evidence and what counts as evidence in their opinion. Both proponents and opponents referred to a variety of evidence in support of their respective positions (Table 5.3). There was an emphasis on peer reviewed papers on both sides, although proponents distinctly emphasized peer reviewed papers as their first choice of evidence. Surprisingly, participants on both sides also relied on evidence originating from the industry and developers of GM crops and the regulatory body (GEAC). The government was also seen as an important source of evidence by both sides. Opponents cited specific scientists

126

more frequently when discussing evidence whereas proponents did not cite specific scientists or studies. Proponents singled out the research of Gilles-Éric Séralini to discredit evidence from opponents. His paper was also cited by the Environment Minister when imposing the moratorium on Bt eggplant. Séralini continues to remain popular in India and was mentioned multiple times by opponents. Some proponents mentioned their own research as evidence.

127

Table 5. 3. Interview Data – Sources of Evidence

Pro-Bt Eggplant Anti-Bt Eggplant Expectation: Proponents and opponents provide evidence to support their position (Herring, 2015). Results: There was an emphasis on peer reviewed papers on both sides. Interviews participants also relied on evidence originating from the industry and developers of GM crops. The regulatory body (GEAC) and the government were also seen as important sources of evidence by both sides. Expectation: Proponents privilege peer -reviewed Expectation: Opponents highlight other sources scientific evidence of evidence in addition to peer -reviewed scientific evidence Results: Proponents privileged peer -reviewed Results: Opponents mentioned peer reviewed scientific evidence papers but also cited evidence from specific scientists Relevant Quotations Peer-Reviewed Papers: Peer-Reviewed Papers: I get my information first from the published Judy Carmen’s paper on pig feeding studies which papers, these all are peer reviewed papers. (Pro- is one of the best papers that I have read on large 01) panels (Anti-04)

I always rely on peer-reviewed journals look at ecological alternatives from peer-reviewed publications … my basis fundamentally by the papers etc. (Anti-06) peer-reviewed journals and publications (Pro-06)

Peer-reviewed, I go by peer-reviewed but then I also read up… (Pro-11)

Industry/Developers: Industry/Developers: so the evidence is in the data that is submitted by may be some parts of dossiers that crop the developers (Pro-07) developers submit to regulators (Anti-08)

Regulatory Body/Government: Regulatory Body/Government: I get my information from the government minutes of meeting of the regulatory body. … agencies who approve the product. (Pro-01) correspondence between governments … official documents (Anti-08)

Scientists: Scientists: People who actually work in the field, who know At the time, they took a Cry1Ac gene and put it much more than me (Pro-09) into Brinjal, low quality genetics, David Andow … forecast that this is, it’s a bad design, low impact toxic, you will have resistance in four to 12 years and Bt Brinjal will fail (Anti-05)

Dr. PM Bhargava who was the Supreme Court nominee to the GEAC and he has headed the … Center for Cellular and Molecular Biology … he himself said the number of tests results were inadequate (Anti-06)

128

In the course of the interviews a more nuanced understanding of evidence began to emerge. Authority and credibility of certain sources were questioned:

So, I don’t want to kind of take anything which is from a media report … media report there was … thousands of sheep died and 10,000 cattle died … That doesn’t hold value for me (Pro-06)

I ignore anything that comes from the industry, anything. I ignore Cornell University because I've done the research, I ignore if there is a scientist who says something that goes diametrically against independent science which I respect (Anti-05)

the problem is that you are not allowing the evidence to come out in … peer-reviewed scientific publications by simply suppressing research like this … you cannot simply say it is not in peer- reviewed journal, then … there is no evidence for this because this suppression is going on (Anti- 09)

Proponents also raised questions about the evidence put forth by opponents:

everyone cherry picks data. … they will quote that … French-Italian guy [Gilles-Éric Séralini] … his paper, got retracted … but they will quote that paper and they can scare the wits out of anybody (Pro-11)

the sad part is the anti-GM people don’t have any specific thing … They just say this may happen, this we heard happens, this may happen 30 years from now. So, these are things that we cannot prove. So, when science cannot prove, it appears like we are at a loss for answers and they have an upper hand over the debate. (Pro-11)

the major contention from activists is that this data [from developers] is not reliable so that is actually where they have to justify why it is not reliable … The developer follows the process. The regulatory agency follows the process … So, when people question the data and question the regulators then you can’t do anything about it. (Pro-07)

Discussion:

Comparing findings from media and interview data, both proponents and opponents cite evidence to support their position though there are some differences in the way they use evidence. Both pro- and anti-Bt eggplant policy narratives in the media contain evidence directly from scientists (in the form of direct quotes/summary/paraphrase) but interviews with other stakeholders indicate that it is opponents who mention individual scientists’ research, while proponents generally more broadly mention the existence of peer reviewed research to support their position. Proponents distinctly articulate their privileging of peer reviewed research over

129

other sources of evidence indicating that it is peer reviewed research that, ultimately, matters.

NPF research shows that “there is no such thing as “just the facts” since” these are always embedded in a narrative story (Schlaufer, 2016, p 108). As illustrated in the quotes above, the interview data reveal a more complex narrative surrounding evidence emerges where the credibility of the sources of evidence is contested. Industry funded research and research originating from regions or institutions known to be supportive of agricultural biotechnology are viewed with skepticism by opponents, while some opponents also claim that there is a lack of peer reviewed research against agricultural biotechnology because it is being suppressed and there is inadequate funding for such research. Proponents blame media reports for providing misleading evidence about the risks from GM crops. Although opponents cited research from multiple scientists to support their claims (David Andow, Judy Carmen, PM Bhargava), proponents mainly singled out the research of Gilles-Éric Séralini, to indicate that scientific evidence provided by opponents has been discredited since Séralini’s paper was retracted. Thus, we see that both proponents and opponents are acutely aware of the importance of evidence to support their positions and, hence provide evidence. But the credibility of evidence remains under contention.

Perception of Risks and Benefits in Agricultural Biotechnology

Findings in the Media Data:

An analysis of risk and benefit information in media coverage (Table 5.4) reveals that both pro- (73.08 percent, n=19) and anti- (80.82 percent, n=59) Bt eggplant policy narratives frequently contain risk and benefit information. While the pro-Bt eggplant narratives highlight economic benefits (68.42 percent, n=13) much more frequently, followed by environmental/health benefits (36.84 percent, n=7), anti-Bt eggplant narratives focus more on

130

environmental/health risk (72.88 percent, n=43), followed by economic risk (23.73 percent, n=14).

Table 5. 4. Risks and Benefits in Media Data (n=99)

Pro-Bt Eggplant (n=26) Anti-Bt Eggplant (n=73) Presence of Risk/Benefit Information 73.08 (19) 80.82 (59) Absence of Risk/Benefit Information 26.92 (7) 19.18 (14) Total 100.00 (26) 100.00 (73) Type of Risks and Benefits Pro (n=19) Anti (n=59) Social Risk 0.00 (0) 11.86 (7) Economic Risk 0.00 (0) 23.73 (14) Environmental/Health Risk 5.26 (1) 72.88 (43) Social Benefit 10.53 (2) 0.00 (0) Economic Benefit 68.42 (13) 5.08 (3) Environmental/Health Benefit 36.84 (7) 0.00 (0) Other 31.58 (6) 32.20 (19)

Findings in the Interview Data:

When interview participants were asked about risk and benefit information, a similar pattern emerges (Tables 5.5 and 5.6). Opponents focused on environmental and health risks both of Bt eggplant in particular and agricultural biotechnology in general. When proponents discussed environmental/health risks, it was in terms of how they can be contained or are non- existent. Similarly, opponents raised economic risk concerns, but these were not raised as much by proponents. Although some proponents acknowledged potential risks, others stressed that there were no risks. Similar to the media coverage, proponents emphasized benefits much more including environmental and health benefits as well as economic and social benefits and de- emphasized the risks. Very few opponents discussed benefits of agricultural biotechnology and only in context of economic benefits while others discussed benefits of banning it.

131

Table 5. 5. Interview Data: Perceptions of Risks

Pro-Bt Eggplant Anti-Bt Eggplant Expectation: Proponents de-emphasize the risks Expectation: Opponents invoke multidimensional risk including environmental, human health, ethical, and religious considerations (Herring and Paarlberg, 2016; Tosun and Schaub, 2017) Results: Although some risks were Results: Opponents invoked multidimensional risk acknowledged, most proponents do not believe including environmental, human health, and there are risks in agricultural biotechnology economic risks Relevant Quotations Risks are Manageable: Environmental and Health Risks: For me one of the fundamental risk[s] … is the there were concerns over health effects (Anti-02) development of resistance of the pest …. So, I would expect anybody who’s going to grow or there’s possible impact on the human health (Anti- who is going to advocate that once if it’s 07) approved, to have a very robust resistance management. (Pro-06) There's no question of the toxicity of the Bt gene in the protein. It is toxic and you’re putting it into Ends Justify Means: food. … the risks I think have to be foremost The higher cost that we pay, would it justify the environmental … No regulator anywhere in the higher production that we get. Those issues are world tests for Endocrine disruption and there is a there. (Pro-04) very strong suspicion more than a suspicion that the EPA and Monsanto have known for 30 years No Risk: that glyphosate is an Endocrine disruptor and I am a firm believer - nothing is going to happen. they’ve hid it (Anti-05) (Pro-03) Economic Risks: I don't think there'll be a negative consequence. 65% of our farming is small holder farming, poor (Pro-04) people. You’re forcing them to buy seed every year (Anti-05) environment safety has been tested, health safety has been tested and people have been consuming 60% to 65% people depending on agriculture or Bt so those concerns I don't think are valid related activities. You cannot jeopardize the very anymore (Pro-09) foundation of the economy. (Anti-06) number of publications have shown that there is no additional adverse impact of any genetically engineered crop on the environment and its safety was never in question (Pro-10)

132

Table 5. 6. Interview Data: Perceptions of Benefits

Pro-Bt Eggplant Anti-Bt Eggplant Expectation: Proponents highlight benefits Expectation: Opponents claim there are no outweigh the risks benefits from agricultural biotechnology Results: Proponents highlighted benefits with no Results: Opponents largely claimed there are no mention of risks benefits from agricultural biotechnology Relevant Quotations Environmental and Health Benefits: Economic benefits: from a simple standpoint of reduced chemical load Ag biotech is one intervention that you have to on the farm (Pro-05) push up your yields. (Anti-02) benefits [are] certainly environmental friendly, No Benefits: pesticide reduction, much quality cleaner products I can't see the benefits from HT and BT crops at for the end users (Pro-06) all. (Anti-05)

Economic Benefits: Benefits from Ban: You will apply less pesticide. and with that the Indian population gains to have straight food healthier crop (Pro-01) and not have -- you know a technology, a techno fix in the name of progress, development, etcetera it’s a much safer product health wise. as a being pushed down that road. It's not a private consumer, I am benefited. farmer he is getting far gain as much as, the country gains to maintain its more productivity. seed company is getting higher biodiversity and maintain its stand that we are not price for the seed. (Pro-03) here as a dumping ground, for technology, which other countries reject. (Anti-06) From the economic standpoint for the farmers there’s a better utilization of input costs because your marketable yield goes up (Pro-05) there would have been yield benefits (Pro-07)

Social Benefits: If you prove that biotech food is safe, then it has a positive impact on food security (Pro-04) there is definitely more social angle to it and it certainly some of these technologies would have resulted in better social development, not just the economics. (Pro-06)

133

Discussion:

Based on the media and interview data, it is evident that the risk discourse is fundamental to agricultural biotechnology. But it is prevalent mainly in the policy narratives of opponents.

Proponents, on the other hand, invoke what may be termed as a ‘benefit discourse,’ which privileges economic benefits as well as environmental and health benefits of agricultural biotechnology over risks. Chong (2005) surveyed eggplant farmers in India and found that farmers are mainly influenced by economic benefits given the uncertainty of their livelihoods in rain-fed farming systems such as India. Perceived benefits have shown to have a higher statistical influence on consumer acceptance (Hamstra, 1995) and perception of benefits may outweigh perception of risks (Gaskell et al., 2004). Demonstrated in the quotes above, proponents are convinced that science has proven that there are no risks to agricultural biotechnology given that environment and health safety have been tested and Bt has been consumed widely invalidating concerns (Pro-09). Opponents are equally convinced of the risks, as shown in the quotes above, with one emphatically claiming that one cannot question the toxicity of the Bt gene in the protein (Anti-05). While proponents continue to espouse economic benefits of agricultural biotechnology, opponents emphasize that it will lead to exploitation of farmers due to rising costs of seeds and dependence on multinational organizations. When imposing the moratorium, the Environment Minister had stated that “Very serious fears have been raised in many quarters on the possibility of Monsanto controlling our food chain if Bt- brinjal is approved” (qtd. in Herring, 2015, pp 170-171). Additionally, he stated that there was a need for “independent scientific studies [to] establish, to the satisfaction of both the public and professionals, the safety of the product from the point of view of its long-term impact on human health and the environment” (qtd. in Gupta, 2011, p 739). Thus, although evidence of risk has

134

been under contention, proponents deny that any such risk exists and instead focus on benefits.

But these are denied by opponents who cannot see any benefits from herbicide-tolerant and insect-resistant crops. Thus, diverse and contested framings of risks and benefits and socioeconomic concerns continue to prevail.

Moral Notions of Risk: Lastly, this study examined moral notions of risk in the context of natural vs. unnatural risk which could only be analyzed through interview data (Table 5.7). Since media articles focused on the policy issue in context of policy core beliefs as opposed to deep core beliefs, it was difficult to consistently code for moral notions of risk in the media data. Interview participants were asked about their thoughts on agricultural biotechnology in general and in context of the relationship between humans and nature. Proponents considered agricultural biotechnology as a natural progression of agriculture and not an unnatural intervention.

Opponents did not believe agricultural biotechnology to be natural.

135

Table 5. 7. Interview Data – Moral Notions of Risk

Pro-Bt Eggplant Anti-Bt Eggplant Expectation: Proponents consider genetic Expectation: Opponents invoke notions of engineering as natural or part of agriculture ‘unnatural risk’ and ‘tampering with nature’ (Sjöberg, 2000) Results: Proponents considered genetic Results: Opponents invoked notions of engineering as natural or part of agriculture ‘unnatural risk’ and ‘tampering with nature’ Relevant Quotations In this case, we are very very specifically it is my understanding that agricultural transferring one or two or three genes depending biotechnology is unnatural, it is not okay (Anti- on the product. … That's not tinkering with 01) nature (Pro-01) when you intervene in that you do expect that in nature everything is modified whether it is things can go wrong (Anti-04) artificially modified by introducing a gene in a particular plant variety and kickstarting the It is a laboratory technology. It cannot be made change. (Pro-04) outside the laboratory, so I don’t see anything that is natural (Anti-05) nature is actually the biggest genetic engineer in a way. You know it’s happening all the time when you're tinkering with genes, in nature, that (Pro-05) can then replicate itself as in the case of agriculture (Anti-06) this is a simple evolutionary step forward … a logical step of improved plant breeding (Pro-06) insert a gene of a living being, in a plant or in a seed you are intervening so that is unnatural I don’t really think there is anything that is (Anti-07) contrary to what nature has been doing (Pro-07)

I think agriculture from the beginning, if we have started, we are not exactly in tune with nature (Pro-11)

136

Discussion:

Research shows that people’s perception of risks is based upon belief systems and that notions of morality dominate risk perception (Sjöberg, 2000). Since humans hold a deep skepticism toward the unnatural, ideas of ‘unnatural risk’ become significant in debates over new technologies. Chong (2005) showed no evidence of moral concern among eggplant farmers in

India. But findings in the present study indicate that moral notions of risk are central to risk perceptions of elite stakeholders in agricultural biotechnology in India. Both proponents and opponents invoke ideas of ‘natural’ and ‘unnatural’ and weave their own narrative of the evolution of agricultural biotechnology. While proponents believe that genetic engineering and agricultural biotechnology are merely next steps in the evolution of agriculture, opponents want to confine it within the laboratory since it is an unnatural intervention.

Conclusion, Limitations, and Future Directions

The goal of this study was to examine policy narratives in agricultural biotechnology policy in India to understand how proponents and opponents of agricultural biotechnology use evidence to support their claims about risks and benefits. It did so by examining the setting and plot in the policy narratives under consideration through an analysis of risk, benefit, and evidence information embedded in policy narratives. Setting and plot remain an underexplored area in NPF research. Setting and plot in policy narratives discuss physical aspects of policy problems and the causal relationships underlying these problems (Lawlor & Crow, 2018). Setting includes the science or evidence about the problem as well as evidence about known or unknown risks (Crow & Jones, 2018; Jones & Radaelli, 2015; Lawlor & Crow, 2018). Furthermore, plot in a policy narrative includes evidence pertaining to human ability to address the policy problem

137

such as existing policies or those that could be implemented to mitigate risk and bring about policy change.

Examining the setting and plot in the Bt eggplant commercialization process revealed that it simply is not just the facts in the story but the way that the facts are narrated through the setting and plot, in terms of credibility of sources and articulation of risks and benefits. We see evidence being “filtered, weighted, and selected” (Herring, 2015, p 160). As shown in previous research, both proponents and opponents relied on evidence, but the current study demonstrates that proponents and opponents questioned the credibility of some sources like media reports and industry funded research. Evidence of risks and benefits is embedded in the narrative story with proponents containing the scope of the issue. They construct a narrative of concentration of benefits largely among the farmers, who are portrayed as the biggest beneficiaries of Bt eggplant commercialization with no mention of risks.

In contrast, opponents expand the scope of the issue and narrate a story of diffusion of costs with adverse health and environmental impacts on consumers, farmers, and the environment portrayed as victims with no mention of benefits. Personal values may influence policy decisions (Mosley & Gibson, 2017; D. A. Stone, 1997) and we find that moral notions of risk play a role in the perception of risks and benefits of agricultural biotechnology. Supporting previous research, opponents highlight multi-dimensional risk (Herring & Paarlberg, 2016;

Tosun & Schaub, 2017) and consider agricultural biotechnology to be unnatural and tampering with the natural process (Sjöberg, 2000) while proponents de-emphasize risks and consider agricultural biotechnology as part of the natural evolution of agriculture. To summarize, an analysis of policy narratives from the Bt eggplant commercialization process revealed that stakeholders use different sources of evidence, question credibility of sources, and proponents

138

de-emphasize risks and exclusively highlight benefits while opponents invoke multi-dimensional risk. Lastly, risk perceptions of stakeholders are influenced by moral notions of whether the risk from the GM crop is natural or unnatural.

The findings from this research are limited to the case study analyzed here and preclude generalizations about how evidence is used in support of claims about risks and benefits in other policy areas characterized by risk. Future research may explore how expectations related to sources of evidence and perception of risks and benefits may differ or align across other policy domains. Additionally, this research examined an understudied area within the NPF, namely setting and plot, based on evidence and risk and benefit information. To develop generalizations on how evidence and risk and benefit information may be used to operationalize plot and setting in the NPF and to develop hypotheses (stemming from the expectations outlined in this paper), it may be useful to apply these variables in other policy domains.

139

References

Andrée, P. (2002). The biopolitics of genetically modified organisms in Canada. Journal of Canadian Studies, 37(3), 162–191.

Bagla, P., & Stone, R. (2013). Scientists clash swords over future of GM food crops in India. Science, 340(May), 539–540. https://doi.org/10.1126/science.340.6132.539

Bandopadhyay, R., Sinha, P., & Chaudhary, B. (2012). Is Bt-brinjal ready for future food?--A critical study. Indian Journal of Biotechnology, 11(2), 238–240.

Baumgartner, F. R., & Jones, B. D. (1993). Agendas and instability in American politics. University of Chicago Press.

Brooks, S. (2005). Biotechnology and the politics of truth: From the green revolution to an evergreen revolution. Sociologia Ruralis, 45(4), 360–379.

Cacciatore, M. A., Scheufele, D. A., & Shaw, B. R. (2012). Labeling renewable energies: how the language surrounding biofuels can influence its public acceptance. Energy Policy, 51, 673–682.

Chong, M. (2005). Perception of the risks and benefits of Bt eggplant by Indian farmers. Journal of Risk Research, 8(7–8), 617–634.

Covello, V. T., Slovic, P., & Von Winterfeldt, D. (1986). Risk communication: a review of the literature. National Emergency Training Center.

Crawley, C. E. (2007). Localized debates of agricultural biotechnology in community newspapers: A quantitative content analysis of media frames and sources. Science Communication, 28(3), 314–346.

Crow, D., & Jones, M. (2018). Narratives as tools for influencing policy change. Policy & Politics, 46(2), 217–234.

Das, K. N., & Bhardwaj, M. (2015, February 21). Modi bets on GM crops for India’s second green revolution. Reuters Publication. New Delhi. Retrieved from http://www.reuters.com/article/2015/02/22/us-india-gmo-insight- idUSKBN0LQ00Z20150222

Gaskell, G., Allum, N., Wagner, W., Kronberger, N., Torgersen, H., Hampel, J., & Bardes, J. (2004). GM foods and the misperception of risk perception. Risk Analysis: An International Journal, 24(1), 185–194.

Golding, D., Krimsky, S., & Plough, A. (1992). Evaluating risk communication: Narrative vs. technical presentations of information about radon. Risk Analysis, 12(1), 27–35.

140

Gupta, A. (2011). An evolving science-society contract in India: The search for legitimacy in anticipatory risk governance. Food Policy, 36(6), 736–741.

Hamstra, A. M. (1995). Consumer Acceptance Model for Food Bioechnology-Final Report. The Hague, The Netherlands.

Herrick, C. B. (2005). ‘Cultures of GM’: discourses of risk and labelling of GMOs in the UK and EU. Area, 37(3), 286–294.

Herring, R. J. (2008). Opposition to transgenic technologies: ideology, interests and collective action frames. Nature Reviews Genetics, 9(6), 458–463.

Herring, R. J. (2014). On risk and regulation: Bt crops in India. GM Crops & Food, 5(3), 204– 209.

Herring, R. J. (2015). Politics of Biotechnology: Ideas, Risk, and Interest in Cases from India. AgBioForum, 18(2), 142–155.

Herring, R., & Paarlberg, R. (2016). The Political Economy of Biotechnology. Annual Review of Resource Economics, 8, 397–416. https://doi.org/10.1146/annurev-resource-100815- 095506

Ho, M.-W. (2000). Genetic Engineering Dream or Nightmare?: Turning the Tide on the Brave New World of Bad Science and Big Business. 2 Rev Upd edition. International Publishing Group Continuum.

Huda, J. (2018). An Examination of Policy Narratives in Agricultural Biotechnology Policy in India. World Affairs, 181(1), 42–68. https://doi.org/10.1177/0043820018783046

ISAAA. (2017). Global status of commercialized biotech/GM crops in 2017: biotech crop

141

adoption surges as economic benefits accumulate in 22 years. Ithaca, NY. Retrieved from https://www.isaaa.org/resources/publications/briefs/53/download/isaaa-brief-53- 2017.pdf

Jasanoff, S. (2005). Designs on Nature: Science and Democracy in Europe and the United States. Princeton, NJ: Princeton University Press.

Jones, B. D. (2003). Bounded rationality and political science: Lessons from public administration and public policy. Journal of Public Administration Research and Theory, 13(4), 395–412.

Jones, M. D. (2014). Communicating Climate Change: Are Stories Better than “Just the Facts”? Policy Studies Journal, 42(4), 644–673. https://doi.org/10.1111/psj.12072

Jones, M. D., & Radaelli, C. M. (2015). The narrative policy framework: Child or Monster? Critical Policy Studies, 9(3), 339–355.

Kalle, J., & Ejnavarzala, H. (2016). Moratorium on Genetically Modified Brinjal in India: Is Evidence-Based Policy making An Adequate Framework? Asian Biotechnology & Development Review, 18(2).

Kellens, W., Terpstra, T., & De Maeyer, P. (2013). Perception and communication of flood risks: a systematic review of empirical research. Risk Analysis: An International Journal, 33(1), 24–49.

Kingdon, J. W. (2003). Agendas, Alternatives, and Public Policies (2nd ed.). Harper Collins Publishers.

Krimsky, S., & Wrubel, R. P. (1996). Agricultural biotechnology and the environment: Science, policy, and social issues (Vol. 13). University of Illinois Press.

Lawlor, A., & Crow, D. (2018). Risk‐Based Policy Narratives. Policy Studies Journal, 46(4), 843–867.

Levidow, L., Carr, S., Wield, D., & Von Schomberg, R. (1997). European biotechnology regulation: framing the risk assessment of a herbicide-tolerant crop. Science, Technology, & Human Values, 22(4), 472–505.

Levidow, L., & Marris, C. (2001). Science and governance in Europe: lessons from the case of agricultural biotechnology. Science and Public Policy, 28(5), 345–360.

McBeth, M. K., Jones, M. D., & Shanahan, E. A. (2014). The Narrative Policy Framework. In P.

142

A. Sabatier & C. Weible (Eds.), Theories of the policy process (3rd ed.). Boulder, CO: Westview Press.

McBeth, M. K., Shanahan, E. A., Hathaway, P. L., Tigert, L. E., & Sampson, L. J. (2010). Buffalo tales: interest group policy stories in Greater Yellowstone. Policy Sciences, 43(4), 391–409.

McInerney, C., Bird, N., & Nucci, M. (2004). The flow of scientific knowledge from lab to the lay public: The case of genetically modified food. Science Communication, 26(1), 44–74.

McMorris, C., Zanocco, C., & Jones, M. (2018). Policy Narratives and Policy Outcomes: An NPF Examination of Oregon’s Ballot Measure 97. Policy Studies Journal, 46(4), 771– 797.

Mehta, D. (2018, December 19). MS Swaminathan and his co-authors are not helping Indian farmers at a moment of crisis. The Print. Retrieved from https://theprint.in/science/ms- swaminathan-and-his-co-authors-are-not-helping-indian-farmers-at-a-moment-of- crisis/165888/

Miles, M. B., & Huberman, M. (1994). Qualitative data analysis: An expanded sourcebook. sage.

Mosley, J. E., & Gibson, K. (2017). Strategic use of evidence in state-level policymaking: matching evidence type to legislative stage. Policy Sciences, 50(4), 697–719.

Nagarsekar, R. (2009, April 6). Biotech to boost brinjal cultivation in Goa. Times of India. Retrieved from https://timesofindia.indiatimes.com/city/goa/Biotech-to-boost-brinjal- cultivation-in-Goa/articleshow/4407541.cms

Nuffield Council on Bioethics. (1999). Genetically Modified Crops: The Ethical and Social issues. Retrieved March 21, 2019, from http://www.nuffieldbioethics.org/publications

Paarlberg, R. L. (2001). The politics of precaution: Genetically modified crops in developing countries. Intl Food Policy Res Inst.

Pew-MacArthur Results First Initiative. (2014). Evidence-Based Policymaking: A Guide for Effective Government. Washington DC.

Pinstrup-Andersen, P., & Schiøler, E. (2003). Seeds of contention: World hunger and the global controversy over GM crops. Intl Food Policy Res Inst.

Raile, E. D., King, H., Shanahan, E. A., McEvoy, J., Izurieta, C., Bergmann, N., … Poole, G. C.

143

(2018). Narrative-based Risk Communication: A Lingua Franca for Natural Hazard Messages? In Midwest Political Science Association.

Ramaswami, B., & Pray, C. E. (2007). India: confronting the challenge–the potential of genetically modified crops for the poor. The Gene Revolution: GM Crops and Unequal Development, 156–174.

Ramesh, J. (2015). Green signals: ecology, growth, and democracy in India. Oxford University Press.

Restall, H. (2014, November 21). Growing a Second Green Revolution. The Wall Street Journal. Los Baños, Philippines. Retrieved from http://www.wsj.com/articles/growing-a-second- green-revolution-1416613158

Rosenau, J. (2012). Science denial: A guide for scientists. Trends in Microbiology, 20(12), 567– 569.

Rubin, H. J., & Rubin, I. S. (2005). Qualitative interviewing: The art of hearing data. Sage. Sato, K. (2007). Meanings of genetically modified food and policy change and persistence: The cases of France, Japan and the United States. Princeton University.

Sato, K. (2013). Genetically modified food in France: symbolic transformation and the policy paradigm shift. Theory and Society, 42(5), 477–507.

Schlaufer, C. (2016). The Narrative Uses of Evidence. Policy Studies Journal, 46(1), 90–118. https://doi.org/10.1111/psj.12174

Schurman, R., & Munro, W. (2006). Ideas, thinkers, and social networks: The process of grievance construction in the anti-genetic engineering movement. Theory and Society, 35(1), 1–38.

Sethi, N. (2009, November 11). Bt Brinjal clearance ignored dissenters? Times of India. Retrieved from https://timesofindia.indiatimes.com/india/BT-brinjal-clearance-ignored- dissenters/articleshow/5216523.cms

144

Shanahan, E. A., McBeth, M. K., Hathaway, P. L., & Arnell, R. J. (2008). Conduit or contributor? The role of media in policy change theory. Policy Sciences, 41(2), 115–138.

Shiva, V. (2000). Stolen harvest: The hijacking of the global food supply. Zed books.

Sjöberg, L. (2000). Factors in risk perception. Risk Analysis, 20(1), 1–12.

Starr, C. (1969). Social benefit versus technological risk. Science, 1232–1238.

Stone, D. A. (1997). Policy paradox: The art of political decision making. London and New York: WW Norton New York.

Tosun, J., & Schaub, S. (2017). Mobilization in the European Public Sphere: The struggle over genetically modified organisms. Review of Policy Research, 34(3), 310–330.

Vembu, V. (2018, December 23). Storm in a scientific teacup. The Hindu.

Wachinger, G., Renn, O., Begg, C., & Kuhlicke, C. (2013). The risk perception paradox— implications for governance and communication of natural hazards. Risk Analysis, 33(6), 1049–1065.

Wangiker, S. D. (2004). Personal interview with Associate Professor of Extension Education, Marathwada Agricultural University, Maharashtra, on February 24, 2004.

Weible, C. M. (2008). Expert-Based Information and Policy Subsystems: A Review and Synthesis. Policy Studies Journal, 36(4), 615–635. https://doi.org/10.1111/j.1541- 0072.2008.00287.x

Weible, C. M., & Schlager, E. (2014). Narrative Policy Framework: Contributions, Limitations, and Recommendations. In M. D. Jones, E. A. Shanahan, & M. K. McBeth (Eds.), The Science of Stories: Applications of the Narrative Policy Framework in Public Policy Analysis (pp. 235–246). New York: Palgrave Macmillan.

Wenzelburger, G., & König, P. D. (2017). Different by Design? Analyzing How Governments Justify GMO Liberalization through the Lens of Strategic Communication. Review of Policy Research, 331–356. https://doi.org/10.1111/ropr.12237

Yin, R. (1994). Case study research. Design and methods. red., Thousand Oaks CA, Sage publications.

Young, K., Ashby, D., Boaz, A., & Grayson, L. (2002). Social science and the evidence-based policy movement. Social Policy and Society, 1(3), 215–224.

145

CHAPTER 6: CONCLUSION

Summary of Major Findings

This dissertation emerged initially with the goal of understanding the role of narratives and, in particular, how the information contained in these narratives pertaining to policy problems, solutions, science, ethics, risk, and other factors is used strategically in the contentious issue of agricultural biotechnology policy in India. Over the course of research, it evolved from an examination of agricultural biotechnology policy in India into an attempt to push the primary theoretical lens, the Narrative Policy Framework (NPF), used in this endeavor forward.

Although the policy decision examined in this study occurred in February 2010, it remains topical due to the ongoing debate surrounding the commercialization of GM crops in

India. At the time that the interviews were conducted in the summer of 2017, genetically modified mustard was being actively considered by the Indian government for commercialization. The policy decision on GM Mustard has been delayed time and again and the moratorium on Bt eggplant is often cited by GMO-opponents as a victory in support of their continued opposition to any transgenic crop including GM Mustard.

Bt eggplant crop developers and government regulators continue to push forward. Eight years after the moratorium, in October 2018, GEAC approached the Indian Council of

Agricultural Research (ICAR) seeking information and data about Bt eggplant from the

Bangladesh Agricultural Research Institute (BARI) where Bt eggplant has been grown since

2013 (Financial Express Online, 2018). At the same time, lawsuits on patenting genetically modified seeds continue to proliferate in India. Over the past two years owing to patenting disputes, the Agriculture Ministry has slashed royalties paid to Monsanto and its joint venture in

India, Mahyco, by local seed companies and the Delhi High Court had ruled that Monsanto could

146

not claim patents in India. This ruling was overturned by India’s Supreme Court in January 2019

(Reuters, 2019). The debate rages on. Needless to say, it is a variety of factors that contribute to the contentiousness surrounding agricultural biotechnology policy in India. This dissertation examined some of those factors through the narratives that are used by stakeholders involved in the debate and, in turn, discovered some distinct characteristics about policy narratives themselves and how they are used by stakeholders.

Chapter 3 examined media coverage from leading English newspapers in India to explore the strategic use of narrative variables in policy narratives. In doing so, it highlighted the important role of incomplete policy narratives in policy debates and outcomes. Policy narratives do not always contain a full suite of narrative components and yet they may be among the most common messages received by the public and political actors. Through an analysis of incomplete narratives, this study further refined the definition of policy narratives and considered which narratives are important from empirical and audience reception perspectives. Recent research shows that the current NPF structure does not take into account several differences that exist among media organizations, platforms, and actors and it remains unclear whether media actively and voluntarily or incidentally participate in coalitions as part of their professional positions

(Crow & Lawlor, 2016). Media as conduit and contributor have been studied (Shanahan et al.,

2008) but by including the category of incomplete policy narratives, one may account for an additional role of the media as disseminator of information without taking sides on a policy issue. This may also account for the differences among policy narratives distributed by advocacy organizations (which have vested interests in promoting their policy preference), on the one hand, and by the media (which may not always promote a policy preference), on the other hand.

147

Results showed that incomplete narratives occur more frequently and contain relevant narrative variables.

One of the other goals of this study was to expand scholarship on the NPF by exploring connections between narrative variables and policy outcomes in a non-U.S. context in order to test the generalizability of this policy process framework through its application in a policy subsystem in a different national and cultural context. Prior meso-level NPF research conducted in this context has revealed challenges from lack of data or accessibility issues for public consumption documents distributed by advocacy organizations (Gupta et al., 2014; Weible et al.,

2016). Hence, this study examined policy narratives available in the media to test the applicability of the NPF and found that narrative variables present across media provide substantial insight into the agricultural biotechnology policy issue in India. There was variation in articulation of problem definitions and characters among coalitions. Thus, in contexts where public consumption documents from advocacy organizations may not be available or easily accessible for meso-level NPF studies, media may prove to be an important and robust source of data. This study attempted to expand the scope of NPF research by using media narratives in a policy and national context not studied before.

Chapter 4 filled an important gap in narrative scholarship through an examination of policy narratives in a different linguistic context not examined before. It investigated the use of narrative elements in policy narratives in agricultural biotechnology policy in India across Hindi and English media coverage and examined the NPF assumption that narratives have generalizable narrative elements irrespective of variation in linguistic context and tested the transportability of narrative elements. The comparison between Hindi and English policy narratives on the same policy issue provided insight into the composition and structure of policy

148

narratives across languages. The narrative elements namely, policy problems, solutions, characters, and evidence were prevalent in the Hindi narratives, which lent support to the transportability of narrative elements and can be used to study issues in policy subsystems in varied linguistic contexts.

The difference in frequency of coverage across Hindi and English media indicated that the issue foci of the two-language media may differ. Since agricultural biotechnology was covered more frequently in the English media, it could indicate that the issue is more prevalent among the elite who are the majority consumers of English media, while the regional language media is more focused on the non-urban public. The difference in issue coverage could also be explained in terms of the difference in interests between the English and Hindi readers and that

India’s Hindi newspapers at the national level employ local journalists to generate news from rural areas to cater to readers interested in local news. Thus, examining English and regional language policy narratives are important to get a more comprehensive sense of the prevalence of particular policy issues at local, regional, and national scales. Thus, its findings validated the transportability of narrative elements in Hindi narratives, indicating variation in the use of narrative elements over time.

Lastly, Chapter 5 dived deeper into the policy issue of agricultural biotechnology through an examination of setting and plot in policy narratives. Operationalizing setting and plot in policy narratives, it focused on the evidence in support of claims about risks and benefits in policy narratives on agricultural biotechnology policy in India and explored moral notions of risk as expressed in actor beliefs. Setting and plot remain an underexplored area in NPF research.

Setting and plot in policy narratives discuss physical aspects of policy problems and the causal relationships underlying these problems (Lawlor & Crow, 2018). Examining the setting and plot

149

in the Bt eggplant commercialization process revealed that it simply is not just the facts in the story but how these facts are discussed and debated among the actors that is relevant in terms of legitimacy of sources and articulation of risks and benefits. We saw evidence being “filtered, weighted, and selected” (Herring, 2015). Both proponents and opponents relied on evidence, but the current study demonstrated that proponents and opponents questioned the legitimacy of some sources like media reports and industry funded research. Findings indicated that stakeholders use different sources of evidence and proponents de-emphasize risks and exclusively highlight benefits while opponents invoke multi-dimensional risk. Lastly, risk perceptions of stakeholders are influenced by moral notions of risk.

A Brief Reflection on the Process

In addition to providing a discussion on the major findings of this discussion, I also include here a few procedural reflections, in terms of research lessons learned. When I formulated this projected, I had planned on collecting public consumption documents and press releases from advocacy organizations, in addition to English media documents and interview data that have been used in this dissertation. But as I began my data collection, stumbling blocks appeared soon enough. Meso-level NPF research conducted on policy issues within the U.S has often relied on public consumption documents that were easily accessible and digitally archived.

But these turned out to be relatively inaccessible in India and not well archived in digital or physical form. But as one door closed, a window of opportunity opened that took my research in a more novel direction. In turn, I expanded my media data to include Hindi media data to test the transportability of the NPF into another language. Keeping an open mind and a willingness to change directions at possible roadblocks were key to maintaining the forward progress of the research.

150

Issues arose in the accessibility of the Hindi media data since Hindi media data are not digitally archived across all media houses in India and, if media data are archived digitally, they are not easily and publicly available. With that realization, began my fieldwork in India. It was a long and arduous process to establish contacts in India, mainly by leveraging existing contacts in the area. I’d never imagined I would find myself physically wading through six years of media data over the course of one sweltering and humid summer in Noida, India. I also spent that summer interviewing stakeholders involved in the process. Given that several of the potential interview participants were high profile (former government ministers, high ranking bureaucrats), I had expected red tape and bureaucracy at every step. Instead, I found myself amazed and humbled by the openness, hospitality, and humility of my interview participants

(and a little too full with all the free chai). In many ways, the interview participants were more accessible and more available than the Hindi media data.

Lastly, I also learnt that conducting good international research can be time consuming, requires a lot of flexibility, and a good deal of funding that is often lacking for graduate students in the social sciences. I was fortunate to have received adequate funding for my fieldwork, but the process took longer than planned and a lot of jugaad, a quintessential Indian term (now

Anglicized and increasingly popularized as a management technique) that the Oxford English dictionary defines as a flexible approach to problem-solving that uses limited resources in an innovative way. I also learned that I find agricultural biotechnology policy research fascinating and that policy positions of stakeholders are more nuanced than would appear and fieldwork was essential to shift my research in that direction. I look forward to applying the lessons I learned here to future research projects.

Areas of Future Research and Closing Thoughts

151

I aim to continue exploring the role of narratives in agricultural biotechnology policy not just in India but through a comparative lens. Each empirical chapter outlined specific areas of future research. Chapter 3 emphasized a need to examine the differences in narratives across different media outlets, which will allow us to understand whether media sources may vary in their role as conduits, contributors, or disseminators in the policy process. Chapter 4 showed that any robust claim on the generalizability of NPF would require extensive comparative NPF studies encompassing several languages. This research only took a first step in that direction in the hope that more comparative linguistic studies in a variety of national contexts will help establish the generalizability of the NPF and its responsiveness to linguistic and cultural specificity in different policy subsystems. The findings from Chapter 5 are limited to the case study analyzed here and preclude generalizations about how evidence is used in support of claims about risks and benefits in other policy areas characterized by risk. Future research may explore how expectations related to sources of evidence and perception of risks and benefits may differ or align across other policy domains. Additionally, this research examined an understudied area within the NPF, namely setting and plot, based on evidence and risk and benefit information. To develop generalizations on how evidence and risk and benefit information may be used to operationalize plot and setting in the NPF and to develop hypotheses, it may be useful to apply these variables in other policy domains.

In a broader sense, further research is needed to understand the influences behind narrative elements in policy narratives such as the cultural context, ideological differences, socio-economic differences. One of the underlying assumptions in majority of NPF research is that narratives matter in influencing policy change. This could be the case in some situations but not in others. There are no definite empirical findings demonstrating that narratives matter in any

152

policy change largely because it is difficult to set controls on other factors (e.g. cultural, ideological, socio-economic) that drive change. For example, in the case study examined in this dissertation, the narratives could be an echo of policy change or the driver of policy change. By examining policy narratives over a significant period of time and across distinct periods of policy change, this dissertation hoped to tackle this issue to some extent. But there is definitely scope for NPF research to develop further in this direction.

Lastly, the issue of agricultural biotechnology policy in India remains close to my heart.

India continues to struggle with food security issues and agricultural biotechnology continues to be cited as one of the means of bringing about a second green revolution to deal with food security problems. Agricultural biotechnology is not a silver bullet to solve the long-entrenched problems in the Indian food systems governance. One of my interview participants stated “you can’t stop the march of technology … don’t close your doors to science … you need to build the public confidence” and I believe that the varied narratives of stakeholders might provide one critical way of deconstructing this debate and will help scholars and decision-makers understand how stakeholder perspectives are informed by their diverse beliefs in science, risk, ethics, as well as by their policy goals. They will provide insight for stakeholders for adoption of best communication practices to influence decision makers and help in the development of better communication practices to aid decision-making. Effective communication strategies are helpful in resolving policy conflicts and in finding common ground from which to move toward constructive policy solutions.

153

References

Financial Express Online. (2018, October 14). GEAC seeks data on GM Bt brinjal from Bangladesh, revives hope for cultivation in India. Financial Express. Retrieved from https://www.financialexpress.com/india-news/geac-seeks-data-on-gm-bt-brinjal-from- bangladesh-revives-hope-for-cultivation-in-india/1348639/

Herring, R. (2015). State science, risk and agricultural biotechnology: Bt cotton to Bt Brinjal in India. Journal of Peasant Studies, 42(1), 159–186.

Lawlor, A., & Crow, D. (2018). Risk‐Based Policy Narratives. Policy Studies Journal, 46(4), 843–867.

Reuters. (2019, January 8). Monsanto patent victory seen spurring biotech investment in India. Reuters. Retrieved from https://www.stltoday.com/business/local/monsanto-patent-victory- seen-spurring-biotech-investment-in-india/article_2c828d10-c76f-55e4-bf3a- 2209f0d18fd5.html

Shanahan, E. A., McBeth, M. K., Hathaway, P. L., & Arnell, R. J. (2008). Conduit or contributor? The role of media in policy change theory. Policy Sciences, 41(2), 115–138.

154

BIBLIOGRAPHY

Andrée, P. (2002). The biopolitics of genetically modified organisms in Canada. Journal of Canadian Studies, 37(3), 162–191.

Arnold, R. D. (1990). The logic of congressional action. New Haven: Yale University Press.

Bagla, P. (2010). After acrimonious debate, India rejects GM eggplant. Science, 327(5967), 767.

Bagla, P., & Stone, R. (2013). Scientists clash swords over future of GM food crops in India. Science, 340(May), 539–540. https://doi.org/10.1126/science.340.6132.539

Bandopadhyay, R., Sinha, P., & Chaudhary, B. (2012). Is Bt-brinjal ready for future food?--A critical study. Indian Journal of Biotechnology, 11(2), 238–240.

Baumgartner, F. R., & Jones, B. D. (1993). Agendas and instability in American politics. University of Chicago Press.

Boykoff, M. T. (2011). Who speaks for the climate?: Making sense of media reporting on climate change. Cambridge University Press.

Brooks, S. (2005). Biotechnology and the politics of truth: From the green revolution to an evergreen revolution. Sociologia Ruralis, 45(4), 360–379.

Cacciatore, M. A., Scheufele, D. A., & Shaw, B. R. (2012). Labeling renewable energies: how the language surrounding biofuels can influence its public acceptance. Energy Policy, 51, 673–682.

Centre For Environment and Education. (2010). National consultations on Bt Brinjal: primer on concerns, issues and prospects. Ahmedabad, India. Retrieved from http://www.ceeindia.org/cee/pdf_files/bt-brinjal-primer.pdf

Chong, M. (2005). Perception of the risks and benefits of Bt eggplant by Indian farmers. Journal of Risk Research, 8(7–8), 617–634.

Covello, V. T., Slovic, P., & Von Winterfeldt, D. (1986). Risk communication: a review of the literature. National Emergency Training Center.

Crawley, C. E. (2007). Localized debates of agricultural biotechnology in community newspapers: A quantitative content analysis of media frames and sources. Science Communication, 28(3), 314–346.

Crow, D. A., & Berggren, J. (2014). Using the Narrative Policy Framework to Understand Stakeholder Strategy and Effectiveness: A Multi-Case Analysis. In M. D. Jones, E. A. Shanahan, & M. K. McBeth (Eds.), The Science of Stories: Applications of the Narrative Policy Framework in Public Policy Analysis (1st ed.). New York: PALGRAVE

155

MACMILLAN LTD. https://doi.org/9781137370129

Crow, D., & Jones, M. (2018). Narratives as tools for influencing policy change. Policy & Politics, 46(2), 217–234.

Damodaran, H., & Sinha, A. (2016, February 8). GM row again, with mustard topping. The Indian Express.

Entman, R. M. (1995). Public Opinion and the Media: How the Media Affect what People Think—and Think They Think. In J. P. Vermeer (Ed.), Media” Res: Readings in Mass Media and American Politics (pp. 55–59). New York: McGraw-Hill.

Fowler, R. (2013). Language in the News: Discourse and Ideology in the Press. Routledge.

Friedlander, P., Jeffrey, R., & Seth, S. (2001). ‘Subliminal Charge’: How Hindi-Language Newspaper Expansion Affects India. Media International Australia Incorporating Culture and Policy, 100(1), 147–165. https://doi.org/10.1177/1329878X0110000114

Ghosh, A., & Boykoff, M. (2018). Framing sustainability and climate change: Interrogating

156

discourses in vernacular and English-language media in Sundarbans, India. Geoforum.

Ghosh, A. K. (2016, April 6). Do we really need GM mustard in India? DownToEarth.

GoI. Constitutional Provisions: Official Language Related Part-17 of The Constitution of India. Department of Official Language, Government of India.