The science and politics of co-benefits : a case study of India’s energy sector

(1)

The

Science

and

Politics

of Co-Benefits:

A Case Study of India’s Energy Sector

MSc Thesis October 2014

International Development Studies (Research) Graduate School of Social Sciences

(2)

(3)

The Science and Politics of Co-Benefits

A Case Study of India’s Energy Sector

MSc Thesis

written by

Jan Paul Mayrhofer (j.p.mayrhofer@gmail.com)

under the supervision of Prof. Joyeeta Gupta and submitted to the Board of Examiners in October 2014 in partial fulfilment of the requirements for the degree MSc in International

Development Studies (research) at the University of Amsterdam.

Date of defence: Members of the thesis committee: November 04, 2014 Prof. Joyeeta Gupta

Dr. Nicky Pouw

“Declaration: I, Jan Mayrhofer, have read and understood the University of Amsterdam plagiarism policy [published on http://www.student.uva.nl/fraudeplagiaat/voorkomen.cfm]. I declare that this assignment is entirely my own work, all sources have been properly acknowledged, and that I have not previously submitted this work, or any version of it, for assessment in any other paper.”

Signature: ... Jan Mayrhofer

(4)

Abstract

Proponents of a more rigorous climate policy have continuously been in quest of new arguments to promote greater action. Examples include no-regrets options or double-dividends to carbon taxation, both alluding to additional benefits of climate policy beyond its merits for greenhouse gas control. One of the more recent framings of this idea is that of ‘co-benefits’, which implies a ‘win-win’ strategy to capture two goals with a single policy measure. The co-benefits concept emerged in scientific circles and has subsequently been consolidated in public policy making. Notwithstanding this genesis, the relationship between the utilisation of the co-benefits idea in both these spheres of activity has not yet been studies. Hence, the question arises:

What is the relationship between the science and the politics of co-benefits?

By using the framework of co-production (as per Jasanoff), this thesis studies the treatment of ‘co-benefits’ in two interrelated but discrete domains, that of science and politics, before drawing conclusions about interaction effects, i.e. the ways in which the processes of knowledge production both shape and are shaped by the wider social and political practices in which they are embedded. In addition, the application of a framework based on discursive institutionalism (as per Schmidt and Hajer) illuminates the significance of both institutional context and discourses in constituting meaning and application of the co-benefits idea. The analysis is based on secondary literature, policy documents and 21 semi-structured interviews with critical stakeholders both inside and outside the Indian policy community identified through a purposive snowballing sampling method. In methodological terms, I apply a literature review, content analysis and the process-tracing method combined with a case study of India in an embedded single-case format analysing a policy programme aimed at transforming India’s energy sector under the heading of a co-benefits approach as an embedded sub-case.

The literature review finds that the evolution of the science of co-benefits is characterized by two processes running parallel to each other: first, increasing diversification and politicisation in the meaning and application of the co-benefits concept and, second, increasing specialization and depoliticisation in the approaches to study co-benefits. In a nutshell, ‘co-benefits’ can be described as a theory that has been monopolized by orthodox economists that have depoliticized the political argument of ‘no-regrets’ into an optimisation approach that largely ignores political realities. I argue that the science of co-benefits has thus reached a paradoxical impasse in which there is a gap between the dynamic nature of the co-benefits concept and the static ways in which it is applied in scientific research.

The case study illustrates the mechanisms through which particular framings of the co-benefits idea, expressed through storylines, came to legitimise certain government policies in a transition to solar energy in the Indian context by means of reconciling competing climate discourses in the country. I will show how the discursive construction of these storylines has been a necessary condition for the adoption of the co-benefits approach given the specific institutional architecture of the Indian political system. It emerges that existing institutions not only impact the political acceptability of storylines, but constitute tangible features of the policy environment which the co-benefits approach must, but thus far failed to, change, hence constraining the realization of multiple benefits upon the implementation of co-benefits policies. In conclusion, this thesis explores the tensions that arise from the application of a depoliticized scientific concept in a highly politicized setting. By means of generalizing the specific mechanisms by which the meaning of co-benefits has been transformed in the Indian case study and comparing these to the scientific usage of co-benefits, it is argued that the co-benefits concept, in both its theoretical and practical application, constitutes a shift in the understanding of climate change that places climate policy in the spectre of the post-political and thereby constitutes to a practice of expert rule and disempowerment. Finally, I will discuss models for remaking the co-benefits approach in both science and politics.

(7)

J.P. Mayrhofer – The Science and Politics of Co-Benefits  iv

List of Figures

Figure 1.1 Environmental Kuznets curve and N-shaped curve ... 2

Figure 1.2 GHG emissions by sector worldwid ... 3

Figure 1.3 The relationship between energy use and GDP growth per-capita for selected countries ... 4

Figure 1.4 Contribution to GHG emissions worldwide in selected periods ... 5

Figure 1.5 Two strands of thought in relationship between the science and politics of climate change ... 7

Figure 1.6 Ontological assumptions of discursive institutionalism ... 10

Figure 1.7 Critical realist view of causation ... 11

Figure 1.8 Procedure for the literature review on co-benefits ... 14

Figure 1.9 Criteria for the selection of co-benefits studies ... 15

Figure 1.10 Mechanisms in the adoption and implementation of co-benefits ... 18

Figure 1.11 Interviewed stakeholders by ‘sector’ ... 21

Figure 2.1 Perspectives on the relationship between science and politics ... 25

Figure 2.2 Ideas in discursive institutionalism ... 32

Figure 2.3 The functioning of discourse in discursive institutionalism ... 34

Figure 2.4 Analytical framework... 37

Figure 3.1 Co-benefits in the literature by definition ... 45

Figure 3.2 Potential co-benefits in various sectors ... 45

Figure 3.3 Total and selected publication on co-benefits in the ScienceDirect database ... 47

Figure 3.4 Publications using co-benefits and related concept in the ScienceDirect database ... 49

Figure 3.5 Scientific nature of publications ... 53

Figure 3.6 Publications classified by applied measurement of co-benefits ... 53

Figure 3.7 Publications classified by sector ... 54

Figure 3.8 Publications classified by scale ... 55

Figure 4.1 India’s economic development (1990-2013) ... 60

Figure 4.2 India’s total CO2 emissions and per-capita C02 emissions in comparison (1990-2010) ... 65

Figure 4.3 India’s energy use and energy imports (1990-2012) ... 67

Figure 4.4 Institutional arrangements created under the NAPCC ... 81

(8)

List of Tables

Table 1.1 Sub-questions of the thesis ... 8

Table 1.2 Coding for quantitative analysis ... 15

Table 1.3 Data sources and justification ... 19

Table 1.4: Validity and reliability in the research design ... 23

Table 2.1 Four schools of institutionalism ... 31

Table 2.2 Operationalisation of key concepts of the analytical framework ... 41

Table 3.1 Three strands of co-benefits ... 43

Table 3.2 Definitions of ‘ancillary benefits’ and ‘co-benefits’ in IPCC reports ... 48

Table 3.3 Comparison of co-benefits and related concepts ... 50

Table 3.4 Appeal and shortcomings of the co-benefits approach ... 55

Table 3.5 Spatial and temporal dimensions of climate mitigation with co-benefits ... 57

Table 4.1 Historical and emerging climate discourses in India ... 64

Table 4.2 Summary of the storylines around the co-benefits approach ... 69

Table 4.3 Example of the co-benefits framework for low carbon strategy ... 74

Table 4.4 Summary of the process of policy construction and communication ... 75

Table 4.5 Division of responsibility between union and state government ... 76

(9)

J.P. Mayrhofer – The Science and Politics of Co-Benefits  vi

List of Acronyms

ACP Asian Co-Benefits Partnership BJP Bharatia Janatha Party

CDM Clean development mechanism CII Confederation of Indian Industry CO2 Carbon dioxide

CoP Conference of Parties CPR Centre for Policy Research CR Critical realism

CSE Centre for Science and Environment DC Developing country

DI Discursive institutionalism

EGLCS Expert Group on Low Carbon Strategies for Inclusive Growth

EU European Union

FICCI Federation of Indian Chambers of Commerce and Industry GDP Gross domestic product

GHG Green house gas GNI Gross national income GoI Government of India

GW Gigawatt

HDI Human development index HI Historical institutionalism IC Industrialized country

IGES Institute of Global Environmental Strategies IMF International Monetary Fund

INC Indian National Congress

IPCC Intergovernmental Panel on Climate Change JNNSM Jawaharlal Nehru National Solar Mission LDC Least developed country

MNRE Ministry of New and Renewable Energy MoEA Ministry of External Affairs

MoEF Ministry of Environment and Forests MoP Ministry of Power

(10)

NAMA Nationally Appropriate Mitigation Action NAPCC National Action Plan on Climate Change NGO Non-governmental organization

NMEEE National Mission for Enhanced Energy Efficiency

OECD Organization for Economic Cooperation and Development PC Planning commission

PM Prime Minister

PMCCC Prime Minister’s Council on Climate Change PPP Purchasing power parity

PPP Public-private partnerships R&D Research and development RCI Rational choice institutionalism

RGGVY Rajiv Gandhi Grameen Vidyutikaran Yojana RPO Renewable Purchase Obligations

Rs Rupees

SAPCC State Action Plan on Climate Change SEA Solar Energy Authority

SECI Solar Energy Corporation of India SERAC Solar Energy Research Advisory Council SERC State Electricity Regulatory Commissions SI Sociological institutionalism

TERI The Energy and Resource Institute UNEP United Nations Environment Programme

UNFCCC United Nations Framework Convention on Climate Change

US United States

WHO World Health Organisation WWF Worlwide Fund for Nature

(11)

J.P. Mayrhofer – The Science and Politics of Co-Benefits  viii

Acknowledgments

My supervisor Professor Joyeeta Gupta, who has supported me to the fullest both intellectually and emotionally throughout the long and sometimes difficult journey that I have gone through with this thesis, deserves my fullest gratitude. Her sharp and goal-oriented approach to academic work (time slicing and satisficing!) has helped me not to go astray and her intellectual rigour has shaped this thesis in myriad ways. Thank you.

I would also like to thank my supervisor Dr. Nicky Pouw for her support and her willingness to function as a second reader at short notice. Moreover, I owe a special thanks to Prof. Lavanya Rajamani and her colleagues at Centre for Policy Research for their guidance during fieldwork in New Delhi.

I am grateful to my colleagues at the University of Amsterdam, espeically Ana De Velde Harsenhorst, Camilla Forti, Gregor Blair and Jeroen van der Heijden, together with whom I have laid the theoretical foundations for this research project. You are great colleagues and even better friends. The same goes for my colleagues at the Research School for Resource Studies for Development. Countless inspiring discussions with you during the PHD training course have functioned as important signposts on the way towards the completion of this thesis. I would also like to thank my colleagues at Iora Ecological Solutions for giving me insights into the working practices of environmental NGOs in India during an internship at the organisation.

A huge thank you also goes to my informants in New Delhi and everyone who assisted in the process of finding incredibly busy people willing to talk to me. Needless to say, this research project would have been impossible without you.

Last but not least, I would have not been able to complete this journey without my family and friends. I am deeply grateful to Sneha Seth and her family for making me feel at home during my whole stay in India. Jai mata di! A special thank you also goes to my parents and my sister. Thank you mum for making me go abroad to the Netherlands, your support throughout my studies and your encouragement and love.

Jan Paul Mayrhofer Amsterdam, October 2014

(12)

(13)

Introduction and Methodology  1

1. Introduction and Methodology

1.1 Introduction

This thesis analyses both the science of the co-benefits concept by means of a literature review and the politics of the co-benefits approach by means of a case study before drawing conclusions about the science-policy nexus. ‘Co-benefits’ here is understood as an idea. It refers to the causal belief that one can achieve two goals with a single policy measure or programme. Ideas shape our understanding of political problems, define goals and strategies to tackle these and serve as a vehicle to communicate interests in the political process (Béland & Cox, 2011). Moreover, the co-benefits concept emerged in scientific circles and has subsequently been consolidated in the realm of environmental politics. Hence, it is indispensible to study the idea of co-benefits within two interrelated but discrete polities or domains.

The first domain is that of science. In order to account for the processes of knowledge production, I draw on the concept of co-production of science and politics, meaning that:

“Scientific knowledge is not a transcendent mirror of reality [but] both embeds and is embedded in social practices, identities, norms, conventions, discourses, instruments and institutions – in short, in all the building blocks of what we term the social [and hence also the political process]” (Jasanoff, 2004: 3)

Consequently, science is conceived as exhibiting a dual relationship with politics. It is neither a straight-forward reflection of a natural reality, nor a concomitant phenomenon determined by political interests. Instead, science and politics are seen as mutually dependent and constitutive.

The second domain is that of politics. I define politics as:

The processes of argument over joint decisions defined by both interests and ideas transmitted through discourse between semi-autonomous agents in an institutional context, the outputs of these processes in the form of policy and institutional

change and the outcomes of these outputs in terms of delivery of goals 1

This definition of politics embraces the importance of meaning making in the political process – the ‘struggle over the best idea’. The chosen case study concerns the adoption of the co-benefits idea in Indian political rhetoric and its implementation in national-level policies through a policy initiative aimed at a transition to solar energy. By means of studying both the science and the politics of co-benefits and, crucially, the relationships between these domains, this thesis contributes to a general theory of co-benefits that should provide valuable insight for people engaged in either one of the spheres under investigation – scientists and policy makers alike.

This introductory chapter provides the problem definition (1.2), the research questions (1.3), the methodology used to answer the research questions (1.4), the focus and limits of the research (1.5) and a summary of the structure of the thesis (1.6).

1_{This definition encompasses and adopts definitions of politics by Lasswell (1950) and Torfing et al.} (2012: 48) and places them into the context of discursive institutionalism (see, e.g., Schmidt, 2008).

(14)

1.2 Background and Problem Definition

Climate change and development are two complex and intertwined issues. The impacts of climate change on both natural and human systems pose a serious threat to development efforts (Rosenzweig et al., 2001; McMichael et al., 2006; IPCC, 2007a; UNDP, 2007; IPCC, 2014a). The following example illustrates this relationship: climate change causes changes in hydrological systems through changing precipitation and melting ice and snow. This leads to an intrusion of saltwater in groundwater through sea level rise as well as flooding in the short-term and low runoff in the long-term. The resulting impacts on human systems are felt in both social and economic terms. Extreme weather events such as floods cause damages that put a strain on economic growth. Lower agricultural productivity and water scarcity impact livelihoods, especially for the poorer segments of society that have less resilience and adaptive capacity. Hence, climate change compounds already existing vulnerabilities and risks.

Conversely, dominant development imperatives stress energy-led growth to fulfil social and economic goals, which leads to an increase in greenhouse gas (GHG) emissions that are responsible for climate change (IPCC, 2007b). The relationship between economic growth and pollution was initially visualized through the environmental Kuznets curve displayed in Figure 1.1 (see, e.g., Cole et al., 1997 for an empirical analysis of GDP per-capita vs. various

environmental variables).2 The curve is valid for local pollutants but not for global pollutants as

empirical findings demonstrate (see, e.g., Dinda, 2004; Stern, 2004). Instead, an N-shaped curve (Figure 1.1) more closely represents the non-linear growth trends of GHGs which steadily increase as a function of growing per-capita GDP (Dinda, 2004). The N-shaped curves illustrate the theoretical and practical impasse of conventional fossil fuel-intensive growth in the face of climate change. This dual relationship is precisely what makes climate change the “defining development issue of our generation” (UNDP, 2007: 1) and accounts for the need to integrate development and climate in both policy and research.

Figure 1.1 Environmental Kuznets (left) and N-shaped curve (right) (Gupta & van der Grijp, 2010: 12-14)

2

The notion ‘environmental Kuznets curve’ is an adaptation of the original Kuznets curve developed by economist Simon Kuznets who empirically demonstrated a U-curved relationship between increased income per-capita (x-axis) and inequality measured through the GINI coefficient (y-axis).

(15)

Energy forms the crucial link in this relationship since its production is the leading cause of anthropogenic climate change and its consumption the governing premise of development. As can be seen in Figure 1.2, energy production is the largest source of GHG generation (34%),

which mainly comes in the form of CO2 from the combustion of fossil fuels for electricity

generation (IPCC, 2014b) followed by forestry, agriculture and land use change (24%), industrial manufacturing (21%), transport (14%) and buildings (7%). At the same time, energy can be seen as principal component of development in both a metrical and ideological sense. Figure 1.3 shows a strong positive correlation between per-capita gross domestic product (GDP) and per-capita energy use on a country-by-country basis. Moreover, access to electricity and alleviating energy poverty are high on national agendas and in itself often used as one of the statistical indicators for development (see, e.g., World Bank, 2014). Apart from statistical data, ethnographies of everyday life reveal the ideological importance of energy availability. In the Western world electricity has become taken for granted in a way that it only captures our attention when it stops flowing during rare fallouts or when a fuse is blown. The other side of the coin is that roughly 1.3 billion people are without access to electricity and in pursuit of a

light bulb and other basic energy-based technologies.3 Furthermore, a growing middle class in

developing countries is now seeking the comforts of a fridge or car. In order to supply these technologies to a growing number of people, energy hunger of both producers and consumers has to be satisfied. It follows naturally that energy systems are in need of transformation to be able to provide access to a growing population and meet the energy needs of industry in a sustainable way. Energy policy – read: climate and development policy – is a crucial arena in providing these much needed changes.

Figure 1.2 GHG emissions by sector worldwide (2010) (IEA, 2013)

3

Different estimates exist. 1.3 billion is an estimate by the International Energy Agency (IEA, 2012). It should be noted that the number of people who do not have access to reliable 24/7 electricity is much higher. 34% 24% 21% 14% 7% Energy production Agriculture, forestry and land use Industry

Transport

(16)

Figure 1.3 The relationship between energy use and GDP growth per-capita (based on World Bank, 2014)

Despite widespread recognition of the necessity to renew energy systems, governments in both developed countries (DCs) and industrialized countries (ICs) have a poor track record when it comes to decarbonising their economy or fostering transformative changes in energy systems, albeit significant variation between state responses exists (Christoff & Eckersley, 2011). Various reasons account for this paradox. First, in most cases it is simply cheaper to produce energy from non-renewable and carbon-intensive sources. Second, even with falling prices for renewable energy technology, diffusion remains rather slow due to the so-called ‘carbon lock-in’. The term refers to a path-dependency of energy and transport sectors in both ICs and DCs inhibiting efforts to introduce alternative energy technologies due to market and policy failures (Unruh, 2000; Unruh & Carrillo-Hermosilla, 2006). The reasoning is that electricity systems are deeply rooted in socio-economic systems and infrastructures. Structural changes consequentially face technological and political economy constraints, i.e. they are blocked by powerful actors who are profiting from the status-quo. Third, and related, the market system imprisons policy making in the sense that a government’s core imperative is to create and maintain conditions for economic growth (Lindblom, 1982). Even if a government may decide to push forward on the climate issue, it might be punished at the polls by a sluggish economy, a fiscal crisis and capital flight resulting from higher energy prices. Fourth, the benefits of action that accrue from a climate point of view will be felt only in the future and potentially elsewhere, but have to be paid for here and now. Climate change is a global problem that operates in a time-scale that outlasts human lifespan and imagination. It is both an intergenerational (Gardiner, 2006) and an

international (Harris, 2007) collective action problem par excellence.4

4

According to Olson (2009) the logic of collective action functions as follows: “unless the number of individuals in a group is quite small, or unless there is coercion or some other devise to make individuals act in their common interest, rational, self-interested individuals will not act to achieve their common or group interest” (p. 2). Although there are objections to placing climate change in the category of classical collective action problems, both on the grounds of behavioural assumptions (Ostrom, 2000) and time-scale (Gardiner, 2006), collective action theory remains useful to account for inaction on the pressing issue of climate change.

Germany United States Netherlands Australia Argentina India China South Africa Low income countries Middle income countries High income countries Portugal Saudi Arabia 0 1000 2000 3000 4000 5000 6000 7000 8000 0 10000 20000 30000 40000 50000 60000 70000 En ergy u se (k g o f o il e q u iv alent p er cap ita )

(17)

A final point relates to DCs that are facing very particular problems when it comes to climate change and the transformation of energy systems. Their historical contribution to the problem has been rather low. DCs, however, are the most vulnerable to the impacts of climate change (Adger et al., 2003; IPCC, 2007a). Climate policy can impact development by limiting the possibilities of using energy sources. This has been recognized by the international community. The United Nations Framework Convention on Climate Change divides the world into Annex I and Annex II countries (UNFCCC, 1992). The idea was that ICs (Annex I) should provide space for DCs (Annex II) to grow based on the ‘development first approach’ (Davidson et al., 2003; Gupta, 1997; 2014). ICs should reduce their own emissions drastically and, through leapfrog technologies, assist DCs to find an alternative way to a low-carbon society (UNFCCC, 1992). Responsibilities for causing climate change, however, are in the process of shifting towards

Annex II countries (Figure 1.4). In particular, four countries known as BASICs5 are now

developing rapidly and hence increase their contribution to the problem of climate change. Therefore, they have to be part of the solution, which can only be reached through concerted efforts of integration of various development and climate objectives. These countries are facing the double burden of development and climate change – providing access to electricity and meeting the energy needs of their growing industry while keeping emissions in limits. Inducing action is thus particularly challenging in these so-called emerging economies.

Figure 1.4 Contribution to GHG emissions worldwide in selected periods (based on PBL, 2013)

The multidimensionality of climate change and its complex relationships with both development and energy concerns make climate change a highly political issue. There is no blueprint for sustainable development, no panacea for the maladies of a brown economy and no predefined pathway towards a green economy or society. Hulme (2009) illustrates this point by illuminating the myriad of interests, ideas and reference frames (he elaborates on scientific, religious, psychological, media, philosophical, and political frames) which are brought to bear on the climate change issue to ultimately explain ‘why we disagree about climate change’. The two frames of importance here are the scientific and the political framing of climate change.

5

These countries include China, India, South Africa and Brazil (first letters and reversed reading) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1850-1949 1950-1989 1990-2010 2011-2030 Rest of Annex II South Africa Brazil India China Rest of Annex I EU27 USA

(18)

Science had a crucial role to play in both shaping and participating in the climate debate right from the start. Concerns about the potentially dangerous effect of global warming began among natural scientists in the 1930s and intensified from 1970 onwards (Weart, 2010). In the 1990s, social scientists became increasingly interested in the phenomenon of climate change, although not the same degree as their colleagues in the natural sciences (Goodall, 2008). From 2000 onwards, scientific consensuses on the reality of climate change, its anthropogenic causes, and near consensus on the severity of potential harms emerged (Oreskes, 2004). Influential reviews conducted by economists like Stern (2006) and Garnaut (2008) quantified the harm caused by climate change and conclude that benefits of immediate action would outweigh the costs. Reviewing the scientific evolution of climate change, Jasanoff (2010) shows that climate science has produced a powerful narrative about climate change as a global phenomenon requiring urgent political action.

A scientific consensus on climate change, however, does not mean that there is consensus on political action. Conversely, the politics of climate change are more bitter than ever before due to various reasons outlined before. This can perhaps best be observed in the international arena at consecutive summits. Various editions of this ‘tragicomedy’ in successive years have shown that, instead of meaningful action, a ‘blame game’ is being played. Indeed, global climate negotiations are in deadlock. Japan, New Zealand and Russia have joined the US in not participating in the second Kyoto period (Gupta, 2014).

Both science and politics are at the heat of climate change. Therefore, the relationship between the two spheres has been studied intensively in relation to climate change. In general terms, the debate seems to evolve between two camps. The fault line between these camps revolves around

the role of science in the politics of climate change.6 On the one hand, several authors (see, e.g.,

Sarewitz & Pielke, 2000; Pielke, 2005; 2007; Nordhaus & Shellenberger, 2007; Metz, 2009; Victor, 2011) have argued for a depoliticization of science. Various lobbyists and politicians like US president Barack Obama have articulated similar concerns (Restuccia, 2013). From this perspective, the key problem is that the existing scientific knowledge is not compelling enough to induce action due to the politicization of science. Therefore, debates about climate action are dealt with at the level of scientific evidence that is used instrumentally to underscore a party’s desired course of action and thus function as a proxy for political debates (Pielke, 2007). The solution is perceived to rest in objective science that can function as a ‘neutral’ tool for guiding policy decisions. This trend can be observed in various assessment reports of the Intergovernmental Panel on Climate Change (IPCC) that presents its results as politically neutral facts based on peer-reviewed literature entailed in its mandate of being ‘policy relevant but nor prescriptive’ (Watson, 2005). The underlying assumption is that the relationship between science and politics is unidirectional (see 2.2.1). Hence, in the problem framing ‘power speaks to truth’ and in the solution framing ‘truth speaks to power’ (Figure 1.5).

On the other hand, several authors (see, e.g., Badiou, 2008; Hulme, 2009; Beck, 2010; Swyngedouw, 2010; 2013) have argued for a (re-)politicization of science. From this perspective, the solution described above becomes the key problem. Swyngedouw (2013) speaks of the ‘non-political politics of climate change’ caused by an all-embracing consensus

6

Science here refers in particular to the role of the social and economic sciences in providing responses to the climate problem in the forms of mitigation and adaptation. As noted before, the debate in the natural science is largely settled.

(19)

about how to deal with climate change. Although dispute is still possible, the debate operates within a paradigm of elite consensus that is subordinated to a techno-managerial discourse (Beck, 2010). This strand of thought insists on a ‘proper’ politicization of climate change through changing the given socio-environmental ordering. Such a shift, however, can only occur if the ‘post-political’ condition, i.e. the predominance of consensual or technocratic representations vis-à-vis conflicting and power-ridden representations of issues such as climate change, can be overcome (Žižek, 1999; Rancière, 2004; Mouffe, 2005). From this perspective, science should engage in a normative debate about climate change that does not obscure underlying discourses, actors, interests, and so on. It is assumed that such a transformative approach to science should enable third-order change in the politics of climate change (Hall,

2011).7 The underlying assumption is that science and politics co-produce the socio-political

order (see 2.2.2) and a depoliticized science leads to a depoliticized politics and a politicised science to a politicized politics, respectively (Figure 1.5).

Figure 1.5 Two strands of thought in the relationship between the science and politics of climate change

The relationship between science and politics in relation to climate change has been discussed by taking policy documents (see, e.g., Beck, 2011) or international organizations (Watson, 2005) as a starting point. The majority of publications discuss the science-policy interface on a general level (see, e.g., Pielke, 2005; Beck, 2010). Fewer articles have taken a particular policy approach or policy idea as a starting point. Therefore, this thesis provides a new perspective on the issue by means of analyzing the benefits approach in both science and politics. The co-benefits approach is optimally suited for this endeavour because it emerged as a relatively abstract scientific concept but is nowadays used by politicians around the world to create consensus around the climate issue and legitimise policy (see 3.3). Moreover, the politics of the co-benefits approach is an understudied area and the study of the relationship to the science of co-benefits is entirely missing (see 3.4). Empirically developing and advancing a general theory of co-benefits thus means advancing knowledge on the science-policy interface of climate change and the politicization debate of both climate politics and science.

7

Third-order change refers to a policy paradigm shift that “takes place when a new goal hierarchy is adopted by policymakers” (Hall, 2011: 657). In relation to climate change, such new goals might relate to the socio-environmental ordering or the restructuring of society through the adoption of, for instance, resilience rather than economic growth as core priority in policy making.

(20)

1.3 Research Questions and Objectives

In consideration of the above stated problem definition, the main research question of this thesis reads:

WHAT IS THE RELATIONSHIP BETWEEN THE SCIENCE AND THE POLITICS OF CO-BENEFITS?

To answer this question, I will trace the idea of co-benefits from its semantic origin in Western academic circles to its currently dominant use in Asia and its ‘non-semantic’ consequences for India and its citizens. The thesis analyses both the science and the politics of ‘co-benefits’ in individual chapters before answering the main research question by comparing insights from both parts of the analysis in the last chapter of this thesis.

Table 1.1 Sub-questions of the thesis

Science (Chapter 3)

Politics (Chapter 4)

1. What is the science of the co-benefits approach? 2. What are the politics of the co-benefits approach?

(a) What are the elements of the co-benefits concept?

Section 3.2 (a) How has the concept been understood and used in Indian political rhetoric to appeal to emerging discourses?

Section 4.4 & 4.5

(b) How did the concept emerge and how is its application distributed?

Section 3.3 (b) How has the concept been adopted in Indian policy making?

Section 4.6

(c) How do scholars approach the study of co-benefits?

Section 3.4 (c) Which role can be attributed to existing institutions in the adoption of the co-benefits approach?

Section 4.7

(d) What are the appeals and shortcomings of a co-benefits approach?

Section 3.5 (d) Which changes in institutions and policy practices can be ascribed to the adoption of ‘co-benefits’ and what are the impacts of these changes?

(21)

1.4 Methodology

1.4.1 Ontological and Epistemological Position

To answer these questions, this thesis adopts a critical realist position that is based on certain ontological and epistemological assumptions. Critical realism (CR) is a philosophy of social sciences associated with philosopher Roy Bhaskar. It has to be positioned between positivism/empiricism and relativism/interpretivism by proposing “a way of combining a modified naturalism with a recognition of the necessity of interpretive understanding of meaning in social life” (Sayer, 2000: 3). It is indispensable to discuss the assumptions made by

CR because they have important implications for the methodology adopted in this study.8

Conversely, the methodology can only produce both internally and externally valid results, if it is congruent with its prevailing ontology (Hall, 2003). In the following, I will thus discuss the main features of critical realism in relation to the topic at hand and the theories, methods and data used in this thesis.

CR entails certain ontological assumptions that render it a distinct philosophy of social science. Ontology refers to “the character of the world as it actually is” (Hall, 2003: 374). In relation to the topic at hand, it thus refers to the nature and operation of the social and political world including claims about causal relationships within that world. As a starting point, CR acknowledges that the world exists independent of our knowledge of it (Sayer, 2000). As a consequence, we have to distinguish between what Bhaskar (2008) has termed the ‘intransitive’

(what is) and the ‘transitive’ (how it is perceived) dimension of knowledge.9

It follows that our knowledge of the world we study is always fallible, theory-laden and conceptually-mediated (cf. empiricism). That there is no ‘golden pathway’ to the truth, however, does not mean that knowledge is all relative and immune to empirical checks (cf. relativism) (Sayer, 2000). Moreover, knowledge “develops neither wholly continuously, as the steady accumulation of facts within a stable conceptual framework, nor discontinuously, through simultaneous and universal changes in concepts” (Sayer, 1992: 5). Second, CR assumes a stratified ontology. In contrast to empiricism, objects are said to have unobservable qualities referring to their powers and structures (the real). These might be activated in certain conditions resulting in observable phenomena (the actual). The empirical, in turn, concerns the domain of experience (Sayer,

8_{It should be noted that methodology is often mistakenly used to refer to the methods used in a study.} The concept, however, cannot be reduced to methods, but refers to the “research strategy that outlines the way one goes about undertaking a research project” (Howell, 2012: ix) in order to increase confidence in the validity of inferences made about the social and political world.

9_{Sayer (2000) illustrates this distinction: “when theories change (transitive dimension) it does not mean} that what they are about (intransitive dimension) necessarily changes too” (p. 11). He gives the example of a theoretical shift from a flat earth to a round earth theory that is unlikely to have caused a change in the actual shape of the earth. I do not hesitate to add that, at least if the social sciences are concerned, this is, in my opinion, only partly true. Rather, I am in accordance with Nowotny (2005) proposing that “the world changes in alignment with the descriptions and explanations provided at particular periods and places over time” (p. 29). Whereas this does not contradict the analytical distinction or the consequences outlined in this section, it gives rise to the need for both an analysis of how (scientific) knowledge impacts the social and political world and reflexivity on part of the researcher, i.e. the systematic awareness creation of how one interacts with the objects of study (see 1.5).

(22)

2000). Third, the phenomena that are dealt with in this thesis (climate change, transformation of energy systems, political and institutional change, etc.) are complex open-systemic phenomena in the sense that they are “generated not by one, but by a multiplicity of causal structures, mechanisms, processes, or fields” (Bhaskar, 2010: 4). In contrast to some of the natural sciences or sub-disciplines of psychology, the possibility to isolate the objects of research and study them under controlled conditions is lacking.

What follows from these three premises is both the need for abstraction in order to reduce complexity and the need for an interpretive understanding or ‘verstehen’ of empirically observed phenomena and their underlying causes. In order to facilitate this exercise, this thesis relies on discursive institutionalism. Discursive institutionalism assumes a recursive relationship between ideas, interests, institutions and discourse (Figure 1.6) that eventually leads to policy change (see 2.3). Science, in this case, plays the role of constructing and shaping ideas such as that of co-benefits. How these variables interact and yield an effect at any point in time for a given political process remains a matter of empirical investigation and causal inference that will be analyzed by taking the idea of ‘co-benefits’ as an entry point. In a nutshell, CR offers a ‘third way’ between constructivist approaches that deny an ‘extra-discursive’ dimension and positivist approaches that ignore the potentially transformative power of variables such as discourse and ideas (Schmidt, 2012).

Figure 1.6 Ontological assumptions of discursive institutionalism (based on Schmidt, 2008)

The epistemological status that can be claimed for discursive institutionalism is a “framework for analysis within which one can describe and analyze phenomena as well as develop and test theories” (Schmidt, 2005: 250). It is different from a theory because it does not make theoretical claims about politics or the science-policy interface, but rather provides us with an underpinning prism that enables the construction and testing of theories within the analytical framework (Stanley, 2012). In my understanding, discursive institutionalism tells us what to look at (which variables to take into account) but not how to look at it (how these variables are related). There is thus a benefit in applying additional theories to identify the mechanisms that are responsible for an idea to cause policy change. Hence, I develop a theory of co-benefits through a literature review of the science of co-benefits (see Chapter 3) that provides testable hypotheses for the subsequent case study in order to both compare science and politics and avoid the caveats of a grounded theory approach (see, e.g., Wilson & Hutchinson, 1996, for a standard review of

(23)

pitfalls). What is more, in CR these causal mechanisms should not be seen from a ‘sucessionist’ view, i.e. as a model in which putative social laws lead to a regular succession of events

(Bhaskar, 2008).10 Rather, in the open systems dealt with here, “the same causal powers can

produce different outcomes, according to how the conditions for closure are broken” (Sayer, 2000: 15). A CR view of causation is thus concerned with finding and explaining contingent

mechanisms (Figure 1.7).11 Accordingly, causality can only be established through an

interpretative understanding of such causal relationships, rather than simply identifying correlations. This is because CR insist that social phenomena are intrinsically meaningful, i.e., partly constituted through meaning and meaning cannot be measured but has to be understood (Sayer, 2000).

Figure 1.7 Critical realist view of causation (Sayer, 2000: 15)

The ontological and epistemological assumptions of CR in relation to this thesis outlined above are constitutive of the methodology applied in this thesis. First, this thesis relies primarily on a qualitative methodology. In principle, CR endorses a wide range of research methods (Sayer, 2000). Looking at the research question, the applied concepts such as discourse and ideas and CR’s insistence on an interpretive understanding of causal relationships, it becomes clear that only a qualitative approach is adequate for the objects and goals of the study at hand. Second, CR’s take on knowledge accumulation as partially continuous implies the necessity of a literature review (see 1.4.2) to build to the accumulative body of knowledge on the studied topic, but to do so in a critical manner by questioning the assumptions and frameworks

10_{In this positivist tradition, causation is established through finding regularities in repeated trials (Sayer,} 2000).

11_{The figure shows that the world has ontological depth, rather than merely consisting of patterns of} events (cf. ‘successionist’ view). Here “events arise from the working mechanisms which derive from the structures of objects, and they take place within geo-historical contexts” (Sayer, 2000: 15). For example, a market (structure) explains how the price for a good is determined (effect) through the market mechanisms of supply and demand. This does not happen in the same way for all goods and all geo-historical contexts (as neoclassical economists working in a positivistic, i.e. ‘succesionist’, view of causality would claim). Much depends on other conditions that are determined by the geo-historical context such as the information actors in the market have and government intervention (‘there is no such thing as a free market’, see, e.g., Chang, 2010: 1-11).

(24)

underlying previous studies.12 Third, CR’s epistemology implies assumptions about the generalisability of knowledge claims. Due to the multi-causal and contingency-based nature of phenomena, we are unable to generalise phenomena beyond their spatio-temporal boundaries in which they are observed (Sayer, 2000). It is, however, possible to make contingent generalizations, i.e. claims about how (general) mechanisms work under specified conditions (George & Bennett, 2005). The method of choice is a case study enabling the theorizing about the political processes (=mechanisms) beyond their spatio-temporal specificity (see 1.4.3). Fourth, the case study has to be combined with the method of process-tracing. Process-tracing is a valuable tool to achieve greater systemization in the use of qualitative methods to identify causal mechanisms in a complex process (Collier, 2011). Process-tracing is optimally suited for research from a CR point of view (see 1.4.4). Fifth, assumptions made by CR also have an impact on both data collection (see 1.4.5) and data analysis (see 1.4.6). The single most important factor here is the interpretative dimension underlying CR social science research that

results in a double hermeneutic (Giddens, 1982).13 This gives rise to potential limitations and

the need for reflexivity (see 1.5). Sixth, due to a distinction between a transitive and an intransitive dimension of knowledge, the transitive dimension of knowledge production, and hence science, becomes an object of study in its own right (see 2.2). Finally, and perhaps most importantly, CR bears an intrinsic relationship to critical social research. Bhaskar (2008) himself points to the emancipatory potential of CR. It involves looking behind the appearances of the world into (potentially false) shared meanings that underpin action resulting in social problems (Sayer, 2000). To disclose actions as being informed by false understandings is to imply that both beliefs and actions ought to be changed.

1.4.2 Literature Review

The literature reviews conducted in Chapter 3 concerns the science of co-benefits. It ultimately serves three purposes: first, to critically analyze current knowledge on the co-benefits concept by means of summary, classification and comparison; second, to identify gaps in knowledge as well as to point out strength and weaknesses of existing literature; third, to use the existing knowledge in the field to develop a theory of co-benefits that can be compared to the use of the concept in social and political practices. The review of scientific literature on the co-benefits concept represents a special case insofar as it is subject of analysis for this thesis rather than merely assisting the analysis by identifying a gap in knowledge. It has therefore been reviewed in a more holistic and systematic way as described below.

12_{Note that this is a middle ground between, on the one hand, radical empiricism (knowledge as facts) in} which the underlying assumptions of previous studies do not have to be questioned because

everything observed is taken at face-value and, on the other hand, radical relativism in which the value of a literature review is undermined by the stance on knowledge as relative to the individual values and cultural background of the researcher, which, in turn, renders knowledge accumulation in the classical sense practically impossible – the same study should always produce different results. 13

A double hermeneutic in the social sciences implies that a researcher does not only interpret the world (as in the natural sciences) but the researched necessarily do so as well. The baseline is that there is a two-way interaction between the researcher and researched in the epistemic encounters. The

researcher is not only influenced by the researched, but conversely the concepts used by social scientists in an interview might in fact change the interpretation of the researched (Giddens, 1982). This might be true in some cases and not in others. In any case, the double hermeneutic leads to the necessity of a two-step interpretation process or as Jackson (2006) elegantly put it: ‘making sense of making sense’.

(25)

Despite – or perhaps rather because of – the ambiguity over its meaning and transformability of the concept, the co-benefits approach has found widespread adoption in both international and domestic policy making. Simultaneously, it has gained widespread attention in academic literature. An overview of studies can be found in Pearce (2000), Rübbelke (2002), Pittel & Rübbelke (2008) and Bollen et al. (2009). These reviews have been conducted on the basis of studies that predominantly use macroeconomic models to estimate benefits of a particular policy or in a particular country. Rather than discussing the concept itself, the reviews are mostly limited to a comparison of previous studies’ findings in quantitative terms. Hence, there is a need for a detailed review of existing literature on the concept of co-benefits to draw a picture of key issues and systematically pinpoint and address gaps in knowledge. Instead of asking the question ‘what are benefits’, the review conducted here concerns the question ‘what is co-benefits’.

A comprehensive review of ‘co-benefits’ and ‘ancillary benefits’, as used in the academic

literature, has been conducted according to the procedure displayed in Figure 1.8.14 A search for

the key terms ‘co-benefits’ OR ‘ancillary benefits’ in the ScienceDirect database revealed that there are 2273 articles making use of either term. Next, articles that had either ‘co-benefits’ or ‘ancillary benefits’ in their title, abstract or keywords have been selected for further analysis. The third step consisted of closer examination of the articles’ content. The articles were filtered for suitability for the analysis at hand according to the laws of Boolean logic, meaning that only articles were selected that fulfilled all three criteria (the area in the middle of the Venn diagram displayed in Figure 1.9). It should be noted that the third criterion, i.e. ‘‘co-benefits’ / ‘ancillary benefits’ being a central part of the analysis’, is a somewhat arbitrary criterion that involves a qualitative judgment. Nevertheless, I argue that the resulting sample of 138 articles provides a reasonable basis for further analysis. In the fourth step, the sample was subjected to both qualitative and quantitative analysis. First, the articles were classified according to a set of seven different variables with multiple categories. The choice for a quantitative approach is grounded in the belief that the sample adequately reflects the population, i.e. the total number of instances the ‘co-benefits’ approach has been studied within academia, and the need to reduce complexity due to the large sample size. Moreover, it allows for a more objective evaluation and the recognition of general patterns and trends in the application of the concept. Second, all articles were scrutinized in relation to their definition and application of the concept of ‘co-benefits’/ ‘ancillary benefits’ as well as their applied methodology in a qualitative manner in order to get a deeper understanding of the concept. The lack of detail that necessarily hinges upon the quantitative approach is consequently compensated for by triangulation with qualitative methods in a mixed methods design.

14_{‘Co-benefits’ is here and elsewhere used interchangeably with the term ‘ancillary benefits’ (see, e.g.,} IPCC, 2007).

(26)

(27)

Figure 1.9 Criteria for the selection of co-benefits studies

The variables used for coding are summarized in Table 1.2. ‘Date of publication’ is an interval variable referring to the year in which the article has been released. All other variables are

nominal variables. ‘Country studied’ consists of the categories developed countries15

, emerging economies (BASIC countries and Russia), and developing countries (all remaining countries). ‘Scale’ refers to the level of analysis in the case the article consists of a case study. Categories include international, national, sub-national (e.g., states in a federal system), local (e.g., a city or a district), and multiple. ‘Definition of benefits’ refers to the application of the term co-benefits (see 3.2). ‘Co-co-benefits’ refers to the type of secondary goals of the policy measure classified according to its nature. Categories include economic, environmental, social and political & institutional. ‘Scientific nature of article’ refers to the approach taken by the author(s). Selected categories include case study, meta-analysis, literature review, theoretical discussion of concept, and other. Finally, ‘measurement’ can be quantitative, qualitative or monetary. In addition, the category ‘not applicable’ is created in case the article at hand cannot be meaningfully classified in relation a particular to variable.

Table 1.2 Coding for quantitative analysis

Variable

Date of

publication Country _studied Scale

Definition of co-benefits Co-benefits (Scientific) nature of article Measurement C at eg o ri es (Year) Developed country Inter-national Climate

co-benefits Economic Case study Quantitative Emerging

economy Country

Development

co-benefits Environmental Meta-analysis Qualitative Developing

country

Regional

(e.g. state) Co-impact Social

Literature review Monetary Local (e.g. city) Political & institutional Theory Multiple 15

High-income country (GNI per capita > $12,616) is used as a proxy for a developed country here.

Climate change mitigation or adaptation measures studied ‘Co-benefits’ / ‘ancillary benefits’ central part of the analysis ‘Co-benefits’/ ‘ancillary benefits’ specified as a result of these measures

(28)

1.4.3 Case Study

The case study conducted in Chapter 4 concerns the politics of co-benefits. A case study is an “empirical enquiry that investigates a [1] contemporary phenomenon [2] within its real-life context, [3] especially when the boundaries between phenomenon and context are not clearly evident” (Yin, 2009: 13). There are several rationales for choosing a case study approach for this thesis that can be related to the sub-parts of the above-cited definition and is further justified by showing the inappropriateness of other methods for this study. First, the adoption and implementation of a co-benefits approach is an on-going phenomenon and therefore a case study has to be given the edge over historical methods. The primary difference here lies in the possibility of a case study to provide an in-depth and holistic investigation relying on multiple sources of information (Creswell, 2012) such as interviews with ‘eyewitnesses’ (see 1.4.5). Second, studying phenomena within a real-life context implies the impossibility of exerting behavioural control. This is precisely the case here and accounts for the unfeasibility of experimental research. In the same vein, a case study can be distinguished from both ethnographic and action research because an emic perspective and the aim to actively change practices, respectively, are not preconditions, but oftentimes unfeasible (Fischer et al., 2006) as is the case here. The third point is related to conceptual validity in the study of complex phenomena such as the policy processes studied here. According to George & Bennett (2005) a case study allows for high levels of conceptual validity due to comprehensive consideration of contextual factors, which is difficult if not impossible in studies based on quantitative modelling. Fischer (2003) makes a similar point for thick descriptive research in case studies, stating that “the key to explaining how [policy and institutional] change comes about has to be grounded in a detailed contextual examination of the circumstances at play in specific cases” (p. 108). Hence, the complexity of the phenomena studied in this thesis principally rules out quantitative approaches such as survey research. In sum, only a case study is able to provide the empirically rich descriptions of the adoption and implementation of the co-benefits approach by looking at the various contextual factors (institutions, interests, discourses, etc.) that may or may not play a role in the process.

The type of case study chosen to evaluate the policy mechanisms, outputs and outcomes of the

co-benefits approach is an explanatory embedded single case study.16 An embedded single case

study analyses one case but distinguishes one or more sub-cases ‘embedded’ within that case

(Yin, 2009).17 Single case studies have attracted criticism for their inability to generate

generalisable knowledge (see, e.g., Ragin, 1989; Easton, 2010). To counteract a lack of external validity, this thesis draws on discursive institutionalism and incorporates a theory of co-benefits established through a literature review because “the use of theories, in doing case studies, [...] becomes the main vehicle for generalizing the results of the case study” (Yin, 2009: 33). A last point refers back to the ontology and epistemology section (see 1.4.1). From a CR perspective generalisation comes from “identifying the deep processes at work under contingent conditions via particular mechanisms” (Easton, 2010: 126). Yin (2009) also points to the fact that case studies (in contrast to experiments) do not enumerate frequencies and cannot therefore

16

Also referred to as nested case study or (multi-)layered case study (Patton, 2005). 17

A holistic case study with an embedded case (as used here) differs from a multiple case study because in the former the context is different for each case, whereas in the latter the sub-unit is studied within a common context (Baxter & Jack, 2008).

(29)

generalize from a sample to a population, but generalize theories. The goal is thus not to show that the co-benefits approach works the same all over the world or even in every sector of the Indian economy, but to elucidate the mechanisms at work under specified conditions and draw contingent generalizations or so-called middle range theories that focus on subtypes of a general phenomenon (Gerorge & Bennett, 2005). By testing certain propositions about co-benefits derived from the literature against their applicability in a specific geo-historical context, this thesis contributes to the credibility of a general theory of co-benefits and simultaneously advanced the theory’s explanatory depth by illuminating the specific mechanisms at play and conditions for closure that may or may not lead to the postulated characteristics of co-benefits. This links to the goal of this study that is: rather than to explore or merely describe phenomena, the type of this case study is explanatory.

The selection of cases and their delineation as bounded systems is an important aspect of every case study and has to be justified (Eisenhardt, 1989; George & Bennett, 2005). The case analyzed here is India. The sub-case is a policy initiatives launched under the heading of a co-benefits approach and aimed at making in India’s energy sector more sustainable through a transition to solar energy. The unit of analysis in both case and sub-case is the co-benefits approach – as a philosophy, a programme, and specific policies (see 2.3). The time period for analysis is roughly the start of the new millennium marking a turning point in the spread of the co-benefits idea (see 3.3) and autumn/winter 2013 when the interviews were conducted. The time frame of more than a decade is appropriate to analyse policy change (Sabatier, 2007). Stake (1995) distinguishes between intrinsic and instrumental rationales for choosing a case in a single case design. Intrinsic cases are not representative for other cases, nor valuable to understand some generic phenomenon, but used to illustrate a particular problem and are therefore interesting in themselves. Conversely, instrumental cases are in itself of secondary interest, but merely serve as a vehicle to develop, test or refine theory. This distinction is analytically useful but far from mutually exclusive in practice. This thesis transgresses boundaries of case selection strategies by relying on both intrinsic and instrumental reasons. On the one hand, the case of both India and India’s energy sector is intrinsically interesting in relation to the topic at hand because much depends on how populous emerging economies like India approach the integration of environmental and developmental objectives and energy plays the crucial link in that relationship (see 1.2). A case study of the problems and solutions in the Indian context has thus a value on its own by informing policy makers and academics alike. On the other hand, from an instrumental point of view, the case displays a high degree of the phenomenon under investigation (the co-benefits approach), which serves as a pre-requisite for an instrumental choice of cases (Pettigrew, 1990). The co-benefits approach was formally adopted as a national strategy in policy documents (see, e.g. PMCCC, 2008). Moreover, both interview data and secondary literature (see, e.g., Dubash et al., 2013) confirms the importance of the concept in Indian policy making. In that sense, the case is optimally suited to understand the co-benefits approach per se and develop middle-range theories about its mechanisms. In sum, the goal is to learn about both the case (India) and the issue of the case (co-benefits). Having justified the selection of case and sub-cases, it is important to keep in mind that “the case study method is correctly understood as a particular way of defining cases, not a way of analyzing cases or a way of modelling causal relations” (Gerring, 2004: 341). I therefore turn to process-tracing as analytical method, to which the following section is devoted.

(30)

1.4.4 Process-Tracing

Process-tracing is a method that enables the reconstruction of a process to make causal inferences for understanding a complex phenomenon, often used in the context of a case study. It is suited for both testing and developing theories as George and Bennett (2005) explain:

“In process-tracing the researcher examines histories, archival documents, interview transcripts, and other sources to see whether the causal process a theory hypothesizes or implies in a case is in fact evident in the sequence and values of the intervening variables in that case” (p. 6)

“Process-tracing is an indispensable tool for theory testing and theory development not only because it generates numerous observations within a case, but because these observations are linked in particular ways to constitute an explanation of the case” (p. 207)

Process-tracing thus aims to identify a causal chain and the causal mechanisms between independent variables, intervening variables, and dependent variables (Collier, 2011). Theories usually encompass several mechanisms that, in turn, increase the theories credibility by enabling more detailed explanations (Checkel, 2005). However, in an analytical framework such as discursive institutionalism, these mechanisms are underspecified. Process-tracing can also be used to inductively generate new hypothesis within the framework of discursive intuitionalism through the development of middle-range theories (George & Bennett, 2005). One key element is thus the focus on causal mechanisms that may be defined as “recurrent processes linking specified initial conditions and a specific outcome” (Maynitz, 2004: 241). A second key element is the focus on the temporal dimension of causality to draw inferences (Blatter & Blume, 2008).

In this research, two processes are being examined that follow a temporal sequence (Figure 1.10). The entry point for analysis in both cases is the formal adoption of the co-benefits approach in India’s policies. The first process takes these as dependent variable and seeks to identify the independent variables that caused the uptake. The second process takes these as independent variable and seeks to identify the dependent variables, i.e. outcomes. Hence, a holistic picture of the co-benefits approach in India can be drawn.

The science and politics of co-benefits : a case study of India’s energy sector

The

Science

and

Politics

of Co-Benefits:

A Case Study of India’s Energy Sector

The Science and Politics of Co-Benefits

A Case Study of India’s Energy Sector

Table of Contents

Abstract

List of Figures

List of Tables

List of Acronyms

Acknowledgments

1. Introduction and Methodology

1.1 Introduction

1.2 Background and Problem Definition

1.3 Research Questions and Objectives

Science (Chapter 3)

Politics (Chapter 4)

1.4 Methodology

1.4.1 Ontological and Epistemological Position

1.4.2 Literature Review

1.4.3 Case Study

1.4.4 Process-Tracing