Governing Forests and Carbon: Reducing Emissions from Deforestation and Forest Degradation (REDD+) in Indonesia

Transcription

1 Governing Forests and Carbon: Reducing Emissions from Deforestation and Forest Degradation (REDD+) in Indonesia Henry Boer BA (Sydney) MEnvSc (Griffith) School of Government and International Relations Griffith Business School Griffith University Submitted in fulfilment of the requirements of the degree of Doctor of Philosophy October 2014

2

3 Abstract Since 2005, negotiations under the United Nations Framework Convention on Climate Change (UNFCCC) have focused on deforestation in developing countries. These negotiations led to the global initiative Reducing Emissions from Deforestation and Forest Degradation (REDD+), an incentive based program to provide finance and technical support to developing countries for mitigation initiatives in the Agriculture, Forestry and Other Land Use sector. REDD+ has rapidly evolved into a complex array of policy reforms, initiatives and demonstration projects in approximately 60 countries across Asia, Latin America and Africa, funded primarily from multilateral climate funds and bilateral development assistance. Under UNFCCC rules and funding agreements, participating countries are required to develop national REDD+ strategies and actions, design funding and benefit distribution mechanisms, and develop robust measurement, monitoring, reporting, and verification (MRV) systems. In addition, governments should implement social and environmental safeguards, such as engaging with local communities and resolving tenure issues, improving governance in the forest sector, and protecting biodiversity. However, the international REDD+ framework is incomplete and it remains at the discretion of participating countries to initiate policies and activities that address local drivers of deforestation, and meet domestic priorities. Indonesia is an early leader on REDD+ and has made a globally significant commitment to reduce emissions and high rates of deforestation. The Government of Indonesia, together with multilateral and bilateral partners, is implementing a REDD+ National Strategy and has initiated a number of administrative and technical capacity building programs, and regulatory and market based policy reforms. The country is also host to over 40 demonstration projects that are trialing different approaches to manage forest carbon, and to provide livelihoods for impoverished local communities. The wide range of policies and measures at the national and sub-national scale is leading to a degree of variance, complexity and also ambiguity in the reform agenda. Overall, implementing major climate change initiatives in the forestry and agriculture sectors remains a problematic undertaking. Governments at all levels struggle with institutional corruption and capacity issues, coupled with broader challenges of democratisation and decentralisation. Global demand for palm oil and pulp and paper is growing, and these sectors provide revenue to regional governments and local employment and investment opportunities.

4 There is a rapidly expanding body of literature on REDD+ that focuses on specific policy areas and project activities across multiple countries, or as evaluations of the status of reforms in particular countries. Research is also attending to the governance of REDD+, with studies addressing the architecture or design of national and subnational programs, the agency of state and non-state actors, and the relationship between discourse, power and institution formation. This focus on specific components of REDD+ means that few studies adopt an holistic approach and analyse the combination of policy reforms and technical initiatives, and importantly how these operate as a broader regime to govern forest resources, land and people. This thesis thus addresses two research gaps and related questions. The first is to analyse how the multiple and overlapping series of policies and practices operate to change or transform the way forests and land-use are governed. The second is to design a suitable theoretical framework to analyse the combination of REDD+ policy reforms. To undertake this analytical task the thesis develops a theoretical framework by combining governmentality theory with three theories used in environmental policy and politics namely, administrative rationalism, neo-liberal conservation and deliberative democracy. The framework applies key concepts from governmentality to each environmental theory including the governing rationality and associated technologies, the formation of subjectivities, and the application of knowledge and discourse. Further, the framework applies the governmentality theory of liberalism, welfare liberalism, neo-liberalism or advanced liberalism, and bio-politics, and utilises theory on the productive nature of power and agency. This combined theoretical framework is then applied to the case study of REDD+ in Indonesia and covers key policy reforms and initiatives across the national, provincial and local scale. The case study also evaluates the major barriers or limitations to the introduction of these reforms. The case study on Indonesia is significant because it provides the first application of the combined theoretical framework to REDD+. To undertake the analysis a rich body of data was collected using qualitative methods including document sources and semistructured interviews with respondents engaged in REDD+ in Indonesia. The thesis argues that REDD+ operates as a governing regime that employs a hybridised intersection of rationalities, technologies and associated knowledge systems designed to change the way forests are managed and utilised. The thesis explores how these rationalities construct the various domains of governing under REDD+, and how contrasting policies and processes attempt to create new subjectivities that turn governments, communities and businesses into sustainable forest managers and ii

5 carbon entrepreneurs. The thesis also outlines a series of overlapping governmental technologies that are designed to control, steer, engage and empower this broad constituency of interests. More specifically, the reforms to land-use administration and increased technical capacity are an attempt by the GoI and partners to create a functioning government that is capable of regulating, controlling and monitoring access to forests. The trialling of carbon markets represents a contemporary neo-liberal approach to steer stakeholder behaviour through the creation of a new and sustainable environmental commodity. In contrast, incentive payment schemes operate as a state managed environmental and welfare program that seeks to change practices by offering development benefits to local communities. The stakeholder participation strategies widely employed in Indonesia are an attempt to democratise and legitimise forest politics and administration. The major challenge in Indonesia is that these new programs of climate conduct conflict with existing resource regimes and economic interests that govern the access and benefits derived from forest resources. iii

6 iv

7 Table of Contents Abstract... i Tables and Figures... ix List of Abbreviations... xi Acknowledgements... xv Statement of Originality... xvii Chapter 1: Introduction Background Research Problem Gaps Hypotheses and Questions Case Study on Indonesia Theoretical Framework Chapter Outline Conclusion Chapter 2: Climate Change and REDD+ Policy Introduction Deforestation and GHG Emissions Drivers and Causes International and National Climate Frameworks Developing REDD National and Sub-national Initiatives Policies to Reduce Deforestation Markets and Incentives Regulations and Protected Areas Finance and Payments Benefit Distribution Carbon Accounting MRV Systems Social and Environmental Safeguards Stakeholder Participation and Engagement Equity, Livelihoods and Forest Tenure Tenure and Carbon Rights Governance Approaches to REDD Climate Change Governance REDD+ Architecture Agency and Discourse Good Governance Limitations with Governance Approaches Further Research v

8 Chapter 3: Research Methodology Introduction Ontology and Epistemology Qualitative Research Case Study Design Selection of Indonesia Data Collection and Analysis Semi-structured Interviews Documents Ethical Conduct Data Analysis Conclusion Chapter 4: Governmentality and Environmental Policy Introduction Governmentality Government and Power Bio-politics, Liberalism and Neo-liberalism State and Non-state Agents Analytical Framework Rationalities and Technologies Problematisation, Knowledge and Subjectivities Critiques and Challenges Governmentality Compared to Governance Environmental Governmentalities Climate Change Governmentalities Developing an Expanded Theoretical Framework Administrative, Economic and Deliberative Approaches Applying the Theoretical Framework Conclusion Chapter 5: Administration and Regulation Introduction Liberalism and Administrative Government Administrative Rationalities and Technologies Limits and Reforms to Administrative Government Administration of REDD+ in Indonesia Technologies of Control and Accountability Deforestation and Public Administration in Indonesia Indonesia s Forestry Sector Administrative Failure Indonesia s REDD+ Program National Strategy and Initiatives Restructuring the Bureaucracy Policies and Regulations Legislative Reforms and Moratorium Forest Management Units and Spatial Planning Compliance and Enforcement Building the MRV System vi

9 5.7 Sub-national Implementation Devolution of REDD Pilot Provinces and Demonstration Projects Challenges to Administrative Reforms Limited Technical Capacity Conclusion Chapter 6: Markets and Incentives Introduction Neo-liberalism and Welfare Liberalism Economic Rationalities, Technologies and Subjectivities REDD+ Markets and Incentives Neo-liberalisation of REDD+ in Indonesia? Deforestation and Market Failure in Indonesia Carbon Market and Incentive Instruments Market Legislation Incentive Scheme Design Finance and Costs Public Funds and Private Investment Calculating the Costs Benefit Distribution and Forest Tenure Payment Distribution Mechanisms Rules for Benefit Sharing Carbon Rights and Tenure The KFCP Demonstration Project Challenges to Markets and Incentives Public Institutions Revenue Shortfalls Loss of Rights and Opportunities Conclusion Chapter 7: Deliberation and Engagement Introduction Deliberative Democracy and Advanced Liberalism Deliberative Rationality and Legitimacy Deliberative Subjectivities and Technologies Limits to Deliberative Democracy Participation and Engagement in REDD International Safeguards Regarding Participation Deliberation and REDD+ in Indonesia Democratic Failure in Indonesia s Forest Sector Participation and Engagement National Policy Consultation Free Prior and Informed Consent (FPIC) FPIC Processes in Indonesia Engaging at the Project Scale Central Sulawesi Pilot Program Participatory Processes in Central Sulawesi Deliberation and Non-state Actors vii

10 7.7.1 Engaging NGOs and Community Groups Challenges to Deliberative Approaches Problems with Consultation and FPIC Engagement as Illiberal Program Conclusion Chap 8: Power, State and Agency in REDD Introduction Governmentality and Power Sovereignty, Discipline and Government Power Bio-Politics, Knowledge and Discourse The Role of the State and Non-State Agents Multiple Topologies of Power in REDD Sovereign, Discipline and Repressive Power Sovereignty Over Forests Disciplinary and Coercive Strategies Productive Government Power Markets in Ecosystem Services Green Economy Resistance from Industries, NGOs and Communities Carbon Bio-power Systems of Forest Surveillance Generating and Transferring Knowledge State and Non-State Agents International Organisations and NGOs Conclusion Chapter 9: Conclusions Introduction Research Objectives and Hypotheses REDD+ as a Regime of Practices Reforming and Extending Forest Governing Challenges to Designing and Implementing REDD Limits of Liberal Government Application of the Theoretical Framework Creating Subjects and Generating Knowledge New Categories of Technologies Theoretical and Empirical Limitations Further Research Appendix A: Interview Question Sheet Appendix B: Ethics and Informed Consent Materials Appendix C: Interviewee Code Key and Pseudonyms Appendix D: Coding Frames References viii

11 Tables and Figures Figure 3.1 Embedded Case Study Design 61 Table 3.1 Progress in REDD+ Countries 62 Table 3.2 Interviewees and Organisations 66 Table 3.3 Document Sources 68 Figure 4.1 Governmentality Concepts 85 Figure 4.2 Combined Theoretical Framework 102 Table 5.1 Liberalism and Administrative Government 110 Table 6.1 Neo-liberalism/ Welfare Liberalism and Carbon Markets 160 Table 6.2 Incentives 172 Table 7.1 Deliberative Democracy and Advanced Liberalism 207 Table 7.2 Deliberative Processes 213 Table 9.1 Governmental Technologies 298 ix

12 x

13 List of Abbreviations AFOLU Agriculture, Forestry and Other Land Use sector AMAN Aliansi Masyarakat Adat Nusantara (Indonesian Customary Community Alliance) APL area penggunaan lain (other land use) AMDAL Analisis Mengenai Dampak Lingkungan (Environmental Impact Assessment) AusAID Australian Agency for International Development BAPPEDA Badan Perencana Pembangunan Daerah (Indonesian: Regional body for planning and development) BAPPENAS Badan Perencanaan Pembangunan Nasional (National Development Planning Agency) BAU business as usual BKPRN Badan Koordinasi Penataan Ruang Nasional (National Spatial Planning Coordination Agency) BPN Badan Pertanahan Nasional (National Land Agency) Bupati local or district governors CDM Clean Development Mechanism CIFOR Center for International Forestry Research CoP Conference of the Parties CSO civil society organisation DKN Dewan Kehutanan Nasional (National Forestry Council) DNPI Dewan Nasional Perubahan Iklim (National Climate Change Council) DPR Dewan Perwakilan Rakyat (House of Representatives) EIA environmental impact assessment FAO Food and Agriculture Organisation of the United Nations FCPF Forest Carbon Partnership Facility - World Bank FFI Fauna & Flora International FIP Forest Investment Programme FLEGT Forest Law Enforcement, Governance and Trade FORDA Forest Research and Development Agency (Ministry of Forestry) FPIC free, prior and informed consent GHG greenhouse gas HPH hak pengusahaan hutan (forest concession right) HPK hutan produksi yang dapat dikonversi (convertible production forest) HTI hutan tanaman industri (industrial timber plantation) xi

14 HTR hutan tanaman rakyat (community-based plantation forest) ICCSR Indonesia Climate Change Sectoral Roadmap ICCTF Indonesian Climate Change Trust Fund IFCA Indonesia Forest Climate Alliance IMF International Monetary Fund INCAS Indonesia National Carbon Accounting System IPCC Intergovernmental Panel on Climate Change JI Joint Implementation KPH Kesatuan Pengelolaan Hutan (Forest Management Unit - FMU) KPK Komisi Pemberantasan Korupsi (Corruption Eradication Commission) KFCP Kalimantan Forests and Climate Partnership LoI Letter of Intent (Indonesia Norway) LULUCF Land-use, Land-use Change and Forestry Sector MoE Ministry of Environment (Republic of Indonesia) MoF Ministry of Forestry (Republic of Indonesia) MPRS Majelis Permusyawaratan Rakyat Sementara (Provisional People s Consultative Assembly) MRV measurement, monitoring, reporting and verification NAMAs Nationally Appropriate Mitigation Actions NFI National Forest Inventory NFL New Forestry Law NGO non-government organisation Permenhut Peraturan Menteri Kehutanan (Ministry of Forestry Regulation) PES payment for environmental services RAD-GRK Regional Action Plan to Reduce Greenhouse Gas Emissions RAN-GRK National Action Plan to Reduce Greenhouse Gas Emissions RED Reducing Emissions from Deforestation in Developing Countries REDD+ Reducing Emissions from Deforestation and Forest Degradation REL reference emissions level RPJPN Rencana Pembangunan Jangka Panjang Nasional (Long-term National Development Plan) R-Plan Readiness Plan - World Bank Forest Carbon Partnership Facility R-PP Readiness Preparation Proposal - World Bank Forest Carbon Partnership Facility RTRW Rencana Tata Ruang Wilayah (Spatial Plan) RTRWP Rencana Tata Ruang Wilayah Provinsi (Provincial Spatial Plan) SFM sustainable forest management SIS safeguard information system xii

15 SVLK Sistem Verifikasi Legalitas Kayu (Timber Legality Verification Standards) TLAS Timber Legality Assurance System TNC The Nature Conservancy UKL-UPL Upaya Kelola Lingkungan-Upaya Pemantauan Lingkungan (Environmental Management and Monitoring Document) UKP4 Unit Kerja Presiden bidang Pengawasan dan Pengendalian Pembangunan (President s Work Unit for Supervision and Management of Development) UN United Nations UNEP United Nations Environment Program UNFCCC United Nations Framework Convention on Climate Change UN-REDD United Nations Collaborative Programme on Reducing Emissions from Deforestation and Forest Degradation in Developing Countries VER verified emissions reduction WALHI Indonesian Forum for Environment WWF World Wide Fund for Nature WTO World Trade Organisation xiii

16 xiv

17 Acknowledgements There are a number of people I wish that thank for their support and encouragement throughout this research project. To Lena, thanks so much for your loving support, patience and great insights over the years, and to Pascale for coming into our lives and bringing such happiness. Thanks also to the Taylor family for your good humour and kindness during the challenges of writing. For my supervisors, Dr Giorel Curran and Assoc Prof Robyn Hollander, I am greatly indebted to your professional approach to supervision and for providing high quality, valuable and at times challenging input on all stages of the research process. Your encouragement and thinking has made me pursue this research to the end. There are a number of people who provided invaluable support for my research in Indonesia, in particular Prof Colin Brown and Mangadar Situmorang Phd. Many thanks for providing the opportunity and facilities to conduct research, and the warm welcome. I would also like to extend gratitude to the Tafsir family for your hospitality. This research would not have been possible without the input and insights from all the interviewees, and others with whom I consulted. Many of the interviewees took the time from demanding schedules to discuss the topic. xv

18 xvi

19 Statement of Originality This work has not previously been submitted for a degree or diploma in any university. To the best of my knowledge and belief, the thesis contains no material previously published or written by another person except where due reference is made in the thesis itself. (Signed) Henry Boer xvii

20 xviii

21 Chapter 1: Introduction 1.0 Background Deforestation has remained a controversial and complex area of the international climate negotiations since the origins of the United Nations Framework Convention on Climate Change (UNFCCC) in Although Parties to the Convention recognised deforestation in developing countries as a significant source and sink of emissions, neither the Kyoto Protocol nor the Clean Development Mechanism (CDM) incorporated provisions or incentives to adequately address the problem. In 2005, the Coalition of Rainforest Nations reignited debate by introducing a proposal to reduce emissions from deforestation in developing countries (RED). 1 The proposal focussed on introducing positive incentives for developing countries, such as finance and technical support, for measures that reduce greenhouse gas (GHG) emissions in the Agriculture, Forestry and Other Land Use (AFOLU) sector (Angelsen and McNeil 2012; Pistorius 2012). In subsequent negotiations the program was expanded to Reducing Emissions from Deforestation and Forest Degradation (REDD+), and adopted as a priority global climate change initiative. REDD+ incorporates an array of mitigation activities covering avoided deforestation and forest degradation, and enhancement of forest carbon stocks through reforestation, sustainable forest management and conservation. Broadly, REDD+ aims to establish an economic value for the carbon stored in forests, provide opportunities for countries to reform the forestry and land-use sectors, and to support sustainable development. Industrialised nations are financing different activities, currently through dedicated climate funds and aid programs, and potentially from compliance carbon markets. In return, participating developing countries are implementing policies and demonstration activities to address the drivers of deforestation. Since its inception REDD+ has rapidly expanded into countries in Asia, Africa and Latin America. There are approximately 60 developing countries engaged at various stages, with initiatives ranging from national policy reforms to local forestry projects. negotiations at the UNFCCC introduced a phased approach, with early readiness initiatives focussed on technical capacity building, and the preparation of strategies or action plans, policies and measures (UNFCCC 2010; Corbera and Schroeder 2011; 1 The Coalition of Rainforest Nations led by Papua New Guinea and Costa Rica tabled the RED proposal at the 15 th Conference of the Parties (CoP) to the UNFCCC held in Montreal in The 1

22 UNFCCC 2013). Countries are required to develop robust measurement, monitoring, reporting, and verification (MRV) systems, a reference emissions level (REL), and financial mechanisms to distribute payments and other benefits to stakeholders. Participating countries also need to conduct measures in accordance with social and environmental safeguards, including broad stakeholder engagement, the recognition of rights and tenure, and co-benefits such as biodiversity protection (UNFCCC 2013). Once these frameworks and technical systems are operational, countries are encouraged to transition to the results-based implementation phase, where payments are conditional upon emissions reductions that are fully measured, reported and verified. The majority of these initiatives are conducted by national governments, often in partnership with multilateral or bilateral aid agencies. At the sub-national level, nongovernment organisations (NGOs) and private companies are engaged in demonstration projects and regularly collaborate with communities and local government agencies. REDD+ aims at transformational change to the socio-economic and political conditions that cause deforestation by shifting the focus of forest management to incentive based carbon mitigation (Agrawal et al. 2011; Brockhaus and Angelsen 2012; Gupta et al. 2013; McDermott 2014). Achieving this change requires a series of policy reforms and initiatives that fundamentally alter how state and non-state agents govern forest resources and land. There are, however, formidable challenges to designing and implementing an effective and equitable REDD+ program. The international framework remains incomplete and there are critical issues to resolve, such as rules regarding the trading of REDD+ credits in carbon markets. Incentives will need to provide livelihoods for the millions of people that depend on forests, and compete against growing global and domestic demand for timber and agricultural products. REDD+ policy reforms must also address a range of crosscutting governance problems in the agriculture and forestry sectors (Angelsen 2009; Agrawal et al. 2011; Gupta et al. 2013). Many developing countries are beset by political corruption and weak public sector institutions, posing significant problems for establishing a performance based payment system and reforming forest allocation and management (Karsenty and Ongolo 2012, p.38). At the local scale, carbon project developers face divisive issues around land tenure, the allocation of benefits, and engaging with the multiple stakeholders that compete over forest resources (Larson et al. 2013; Sunderlin et al. 2014). Indonesia is a global leader on REDD+ and the national government has commenced a substantial policy reform program. The Government of Indonesia (GoI) has pledged 2

23 to cut the country s GHG emissions between 26% and 41% by 2030, primarily by reducing high rates of deforestation. Indonesia s ambitious commitment has attracted billions of dollars in finance and support from multilateral agencies and bilateral partners such as Norway. Since 2007, the GoI has released a National REDD+ Strategy and set up a new coordinating agency, introduced a moratorium on forestry concessions, and new legislation to facilitate carbon markets (Wright 2011; Brockhaus et al. 2012; Sloan 2012; Lima 2013). The country is host to multiple demonstration projects, which are trialling different methods to manage and restore forests. However, delivering REDD+ across Indonesia remains a challenging undertaking. Forests, agriculture and mining are vital to national and local economies, and a major source of revenue for governments (Irawan et al. 2013). The GoI is struggling to address entrenched corruption and illegality in the public sector, and the forestry and agricultural industries remain intertwined with political authority at all levels (Barr et al. 2006a; Moeliono et al. 2009; Gellert 2010; McCarthy 2011; Hapsari 2011; Aspinal 2013). REDD+ is a controversial issue in national and local politics, subject to sustained opposition from industry sectors, and from the many local communities who fear impacts on their rights and livelihoods (Luttrell et al. 2014). 1.1 Research Problem REDD+ has evolved into a complicated array of policy initiatives and projects, operating across multiple levels and jurisdictions. Participating countries are required by international rules and funding agreements to implement certain technical programs and safeguard frameworks. But discretion remains with governments to adopt policies that address domestic priorities around deforestation and economic development. In many countries, governments need to integrate new carbon-based programs with existing forest management and land-use planning frameworks. Countries such as Indonesia are using different market and regulatory instruments that utilise contrasting approaches to manage forests and carbon. Government agencies and project developers also apply various engagement and consultation processes, including mechanisms for free prior and informed consent (FPIC). These overlapping REDD+ processes and activities necessarily engage a range of state and non-state interests, leading to complex decision making and contested authority over the reform agenda. This combination has resulted in what a number of authors argue is a high degree of institutional fragmentation, policy ambiguity and incoherence in the programs and 3

24 activities operating across multiple levels and jurisdictions (Corbera and Schroeder 2011; Angelsen and McNeill 2012; Gupta et al. 2013; McDermott 2014). This thesis seeks to understand how the multiple, intersecting and often contrasting series of policy reforms, project activities, engagement processes and technical programs operate to change forest governance. The thesis also aims to evaluate the ways in which different state and non-state agents engage in the reform program, and how they govern the multiple REDD+ initiatives. An analysis of the construction, design and application of these multiple initiatives is important for understanding how forests, land and people are to be governed under a new carbon based regime. This thesis undertakes this analysis by conducting a case study of the operation of REDD+ in Indonesia Gaps There are two critical gaps in the current literature on REDD+ that this thesis aims to address. The first gap relates to the limited number of in-depth empirical studies that analyse the major initiatives that comprise REDD+ in different countries. While there is a rapidly expanding body of research on REDD+, much of it focuses on specific elements of the reform agenda, often as comparative studies across countries or case studies on demonstration projects. For example, there are numerous studies on MRV and GHG accounting frameworks, carbon markets projects, national funding mechanisms, and multiple case studies on tenure, equity and livelihood issues. While research on the specific components is valuable it does not adequately analyse the combination of initiatives and activities operating in particular countries. There is a need for research that provides a more holistic analysis of the various components, and how they contribute to a new carbon based forest governing regime. Research is also required on the design of REDD+ national strategies and sub-national programs and how these integrate or conflict with existing forestry and economic development policies. This research needs to evaluate the different political, legal, economic, social and cultural elements that contribute to deforestation in specific countries, and importantly how these factors influence the design and implementation of REDD+. The second gap in current research relates to suitable theoretical frameworks to analyse REDD+. Current research applies various governance theories to analyse the REDD+ institutional architecture and the expanding role of non-state actors (Corbera and Schroeder 2011; Agrawal et al. 2011; Brockhaus and Angelsen 2012; Brockhaus 4

25 et al. 2013). For example, multi-level approaches focus on mapping actor relationships and interplay between institutions at different scales (Skutch and Van Lake 2009; Mwangi and Wardell 2013; McDermott 2013), paying attention to MRV systems (Korhonen-Kurki et al. 2013), and land tenure or carbon rights (Doherty and Schroeder 2011). Other governance studies evaluate the relative legitimacy and effectiveness of different institutional modes, such as the operation of carbon accounting methodologies (Lederer 2011; Gupta et al. 2012), in addition to incentive and payment mechanisms at the global and national level (Rosendal and Andresen 2011; Vatn and Vedeld 2013). Recent studies also adopt a discursive-institutional approach to analyse relationships between policy design, discourse and power (Arts and Buizer 2009; den Besten et al. 2014), and the capacity of different actor networks to utilise discourse to influence policy frameworks (Angelsen et al. 2012; Luttrell et al. 2014). The application of governance theories to climate change programs such as REDD+ provides a range of valuable insights, but there are critical limitations. Governance theories often fail to adequately address the process of governing; and importantly how different processes unfold in certain contexts, or the multiple practices employed in constructing and managing the climate domain (Okereke et al. 2009; Lövbrand and Stripple 2012). Further, the focus of governance on shifting public and private authority and the relationship to certain institutional modes does not sufficiently account for the multiple ways governing is conducted by state and non-state agents (Sending and Neuman 2006), what methods or tactics are employed, or how power is produced or exercised in different settings. In the case of REDD+, there is a need for suitable theoretical frameworks to understand how this new domain is constructed, and then made operable and governable through the combination of initiatives and practices. Suitable theories are also required to analyse the ways in which contrasting policy instruments steer and direct multiple state and non-state actors towards goals of carbon mitigation and sustainable forest management. Further, there is a need for theories that can evaluate the engagement of state, non-state interests and individuals in REDD+, and how their views, activities and organisational networks shape policy approaches and potential outcomes (Corbera and Schroeder 2011, p.96) Hypotheses and Questions The research has two broader hypotheses and two sets of questions that respond to the problems and research gaps outlined above. 5

26 The first hypothesis and research questions relate to how REDD+ operates as a governing regime. Hypothesis 1: REDD+ incorporates a diverse and complex series of policy initiatives, technical programs and project activities that operate as a regime of practices designed to transform how forests, land and people are governed in Indonesia. The research questions therefore seek to understand how this complex regime of practices is designed and how it operates in Indonesia, including barriers and limitations. The main questions are: How do the various policy initiatives, programs and demonstration projects that comprise REDD+ in Indonesia operate and what are their intended effects? How do these various policy instruments and programs attempt to govern the interaction between the state, non-state stakeholders and forests? How is REDD+ governed in terms of how different actors engage and exercise authority in the design and partial implementation of initiatives and projects? What are the limitations of these different policy initiatives and programs and what are the political and economic barriers to their implementation? The second hypothesis and related research questions focus on developing theory to answer the above questions. Hypothesis 2: A theoretical framework that combines governmentality with theories of administrative rationalism, neo-liberal conservation, and deliberative democracy offers a suitable approach to analyse and interpret the multiple components to REDD+ in Indonesia. The main theoretical research questions include: How can governmentality theory be combined with, and applied to, three environmental theories of administrative rationalism, neo-liberal conservation and deliberative democracy? What does this combined theoretical framework inform us about REDD+? Is the theoretical framework suitable in this application? 6

27 1.2 Case Study on Indonesia This research will address the research questions by conducting a case study on the design and partial implementation of REDD+ in Indonesia. Indonesia was selected because the GoI and partners are conducting a relatively large reform program, providing the opportunity to analyse the various policy and technical components that comprise REDD+ in developing countries. The case study will analyse the major initiatives and programs underway at the national and sub-national level. In addition, the case study will evaluate the challenges and limitations to the introduction of REDD+ in Indonesia; however the analysis does not offer a quantitative evaluation of the effectiveness of different initiatives. Rather, the aim is to situate REDD+ within the broader Indonesian political context, such as the transition to democracy, decentralisation, and the ongoing problems with corruption in the forestry administration. The case study is significant because it offers the first in-depth application of the combined theoretical framework to a climate change program in a developing country Theoretical Framework Evaluating the complex, multi-level and often disparate components to REDD+ in Indonesia requires a suitable framework that can encompass several theoretical approaches. The thesis therefore develops an analytical framework by combining several theories. The thesis employs a macro framework of governmentality theory that focuses on the various practices of political rule or the conduct of conduct, and how power functions or is produced in government (Foucault 2008; Miller and Rose 2008; Dean 2010; Bröckling et al. 2010; Walters 2012). Governmentality analysis combines two broad aspects of governing, namely the rational understanding or thinking about a phenomenon and acting upon the same phenomenon in order to transform it (Gordon 1991; Miller and Rose 2008, p.15; Dean 2010, p.26). The concept of governmentality thus enables systematic scrutiny of the relationship between techniques of power and knowledge, since thought is intimately embedded within governmental practices that are designed to shape and reform social conduct (Bröckling et al. 2010, p.2). Governmentality theory utilises certain rationalities or logics to understand the multiple programs of government, and explore how these are implemented through associated technologies, the construction of subjectivities or identities, and the application of discourse and knowledge (Lemke 2002; Miller and Rose 2008; Dean 2010). In addition, governmentality theory addresses how 7

28 authorities exercise rule and how individuals self-govern within the broader concepts of bio-politics, liberalism, welfare liberalism, neo-liberalism, advanced liberal government, security, discipline and sovereignty. Governmentality thus represents a relatively broad and flexible concept that encompasses different styles of thought and reasoning, the principles and knowledge that they utilise, the governing practices or technologies they consist of and how they are carried out, and their patterns of correlation with other forms of governing (Rose et al. 2006, p.84). This thesis adopts and applies this broader and flexible theoretical orientation to governmentality (Foucault 2007); one that recognises the heterogeneous and often intersecting practices and approaches to governing (Rose et al. 2006; Dean 2010; Walters 2012); and importantly that the programs of government are subject to change and recombination when deployed in different contexts (Collier 2009). To further develop a theoretical framework this thesis combines this relatively flexible governmentality framework with three key environmental theories and related discourses applied to environmental policy and governance namely, administrative rationalism, neo-liberal conservation, and deliberative democracy (Dryzek 2005a; Kronsell and Bäckstrand 2010; Fletcher 2010). This thesis expands and develops each of these environmental theories into an analytical framework by applying the governmentality concepts outlined above (including rationalities, technologies, problematisation, subjectivities, knowledge and discourse). The thesis argues that combining governmentality theory with the three environmental policy theories provides an expansive and more nuanced set of analytical devices with which to explore the multiple components to REDD+. The theoretical framework helps reveal how REDD+ incorporates a hybridised and divergent collection of rationalities, subjectivities and forms of knowledge, that in turn shape policy responses and determine the ways in which forests can be governed. The analysis also shows that REDD+ operates through a combination of government technologies that are designed to control, steer, engage and empower a broad constituency of public and private agents. 8

29 1.3 Chapter Outline Chapter 2 provides an overview of the expanding body of literature on REDD+ and provides a synthesis of research on the main policy and program areas at the international, national and local level. The purpose for the review is to outline and explore the scope of REDD+ across the various technical, administrative, policy and theoretical fields, and to provide a context for the detailed analysis of Indonesia to follow in Chapters 5 through to 8. The review also provides a critique of governance approaches to REDD+. The Chapter is organised according to major policy areas or subjects identified in the literature, commencing with deforestation and GHG emissions. The review then shifts to the international framework and the design of REDD+ programs in developing countries, including finance and payments, technical rules and MRV systems, and the different policy instruments to address deforestation. The next sections review current research on environmental and social safeguards, and the issues of rights and forest tenure, and stakeholder participation. The final section critically reviews the governance literature on REDD+ and climate change, and identifies the gaps this thesis will address. Chapter 3 outlines the research design and methodology and justifies the selection of the case study. The first section of the Chapter introduces the constructivist ontology and qualitative data collection and analysis methods used in the research. Data for the case study was collected from interviews with a range of professional staff working on REDD+ in Indonesia, and from primary and secondary documents. The second section explains and justifies the selection of the single embedded case study on Indonesia. More specifically, a single embedded case study allows a detailed analysis of the multiple REDD+ initiatives, policy reforms and demonstration activities across national and sub-national levels, but within the same political context. The Chapter finishes with an outline of data analysis techniques and research ethics. Chapter 4 develops the theoretical framework. The Chapter introduces the key concepts of governmentality theory and the three environmental policy theories that expand the analytical framework. The first section provides a critique of Foucault s (1979, 2007, 2008) concepts of governmentality and power that are explored through the transition from sovereign rule and bio-politics to modern liberal and neo-liberal government. The second part introduces key theoretical concepts such as governmental rationalities and technologies, problematisation, subjectivities, knowledge and discourse that will be applied to Indonesia. The Chapter also evaluates 9

30 key critiques of the theory and its application to developing countries. The final section reviews research on governmentality analysis in environmental and climate change politics, and outlines why the combined theoretical framework was adopted. Chapter 5 is the first case study on Indonesia and focuses on the administrative, regulatory and technical reforms. The Chapter provides an overview and short history of forest policy in Indonesia and argues that high rates of deforestation are partially attributable to administrative failures. The Chapter combines theories of liberal governmentality with administrative rationalism, focussing on bureaucratic hierarchy, regulations and scientific expertise. This theoretical framework is then applied to Indonesia to evaluate how the GoI and partners are attempting to establish a new and coordinated REDD+ administration, and overhaul forest policy and management. The Chapter evaluates the uptake of MRV and covers the development of sub-national REDD+ and the ongoing issues of administrative decentralisation. The final sections evaluate implementation barriers such as policy integration, and accountability and corruption in the Indonesian public sector. Chapter 6 analyses the design and establishment of a forest carbon market and incentive scheme in Indonesia. To undertake this analysis the Chapter first outlines a theoretical framework that integrates governmentality theories on neo-liberalism and welfare liberalism, with a recent scholarship on neo-liberal conservation. The analysis focuses on how carbon markets create a new environmental commodity in order to change the way public and private agents use and manage forest resources. The Chapter also evaluates the welfare liberalism inherent within incentive schemes, and the role of public institutions in designing and operating various market and payment mechanisms. The Chapter sections cover financing and payment distribution, rules for benefit sharing, tenure and carbon rights, and the application of various accounting methodologies. A detailed assessment is undertaken into the Kalimantan Forest and Climate Partnership, a demonstration activity that aims to deliver income and benefits to rural communities in return for rehabilitating peat forest. Finally, the Chapter assesses issues of equity, and the procedural and distributive barriers to effective implementation. Chapter 7 analyses how the GoI, multilateral organisations and NGOs have employed public consultation, stakeholder participation, and FPIC in Indonesia. The Chapter develops an analytical framework that combines governmentality theories on advanced liberal government with deliberative democracy and its application to environmental 10

31 issues. The premise of deliberative democracy is that increased stakeholder engagement in decision making can improve the procedural qualities and legitimacy of environmental reforms, and enhance the democratic quality of forest governance (Baber and Bartlett 2005; Dryzek 2010). This theoretical framework is then applied to a detailed study of deliberative technologies used to inform the preparation of Indonesia s National REDD+ Strategy and the sub-national pilot program conducted in Central Sulawesi. The following sections explore how deliberative processes create new subjectivities around forests, and integrate multiple communities, businesses and local authorities in the design and operation of projects. Although participatory strategies are widely applied in Indonesia there remain substantial challenges with the delivery and legitimacy of these processes. Chapter 8 shifts to an analysis of the role of state and non-state agents and the nature of power in REDD+. The analysis provides a synthesis of the material presented in the case study on Indonesia, and extends the analysis to other countries conducting REDD+. The Chapter introduces a theoretical approach based on Foucault s concepts of productive power, as exercised through discourse and knowledge. The Chapter analyses how different actor networks produce or employ power though the use of discursive techniques, often to influence the political process and to expand access to economic resources. The first section evaluates sovereign power expressed though the disciplinary regime of law enforcement and surveillance over forests. This is followed with a discussion of productive power and a more detailed study of environmental markets and carbon bio-politics. The last section addresses the function of the state and the agency of public and private actors engaged in the various REDD+ initiatives. Chapter 9 concludes this study, summarising the results and the analytical and theoretical findings. The Chapter reveals how REDD+ creates a new domain of governing through a series of overlapping or hybridised rationalities and technologies and the application of multiple forms of knowledge. These in turn create a series of new governing subjects such as accountable and competent public administrators, the entrepreneurial businesses that commercialise carbon, and the responsible communities that manage forest environments. The Chapter then considers the significant political and economic challenges to successful implementation of REDD+, and the contradictions this creates for a democratising state such as Indonesia. Finally, the Chapter outlines the contribution of the thesis to theory development in the 11

32 field of climate governmentality, identifies several limitations to the theoretical framework, and suggests areas for further research. 1.4 Conclusion REDD+ has developed into a major new field of climate change policy in developing countries. There are multiple intersecting programs and activities underway that seek to transform forest management, and introduce new carbon market and incentives schemes. REDD+ is an ambitious program that aims to deliver emissions reductions, address issues around rights and equity, and offer sustainable economic development for local communities. However, REDD+ faces significant barriers in terms of the capacity of governments to deliver reforms, increasing global demands for timber and agricultural commodities, and conflict over resource access. REDD+ is therefore a highly complex multi-level process that engages a myriad of state, NGO, business and community interests. Although there is an expanding literature on REDD+, there is a need for more research that addresses how the multiple components in combination operate as a new governing regime for forests and people. This thesis seeks to address this gap by undertaking an in-depth case study of Indonesia that covers the major policy and administrative reforms, technical programs, and sub-national activities. The thesis will also evaluate the role of different state and non-state agents in the reform processes. To conduct the analysis the thesis will develop an innovative theoretical framework that combines governmentality with three key environmental policy theories. The next Chapter provides an overview of the literature in the field of REDD+. 12

33 Chapter 2: Climate Change and REDD+ Policy 2.0 Introduction Since the program Reducing Emissions from Deforestation and Forest Degradation (REDD+) was first introduced to the international climate negotiations a large and diverse body of research has emerged covering different initiatives of the reform program. At one end of the scale is statistical modelling of deforestation and greenhouse gas (GHG) emissions, and at the other is constructivist research on the public discourses that shape domestic programs. The aim of this review is to provide an overview and synthesis of the rich literature on REDD+, but with a focus on the main policy, program and technical components across the international, national and project scale. Chapter 1 identified the areas of REDD+ to be covered in the case study on Indonesia, and this review will address each of these different elements to inform the empirical analysis in Chapters, 5, 6, 7 and 8. However, it remains beyond the scope of this Chapter to cover all the related material on REDD+, or to analyse research underway in multiple countries. Importantly, the review will identify the current gaps in the literature that this thesis will address. Current research on REDD+ provides detailed insights into various initiatives or components, and some preliminary analysis of national policy frameworks. However, there are few studies that provide a holistic and comprehensive analysis of the multiple policies and programs operating in particular countries, or the linkages between national and sub-national activities. This Chapter also provides a critique of the different theoretical approaches to analyse the governance of REDD+. This review shows that governance theories often focus on the institutional architecture and design of particular initiatives across levels, and offer evaluations of the input and output legitimacy. However, these analyses provide limited insights into the function of multiple REDD+ policy reforms, or how they contribute to a broader regime of carbon management. Further, current governance approaches analyse the discourses employed by different actors or networks, but need to account for the ways in which state and non-state actors engage as subjects in the governing of forest carbon. The literature on REDD+ is situated in a variety of disciplines and utilises different methodologies and theoretical approaches. To structure the review, material is 13

34 organised according to major policy areas or subjects, and where appropriate is further differentiated according to methodology, such as case studies and comparative analysis. The first section outlines the different drivers of deforestation and GHG emissions in developing countries. The review then shifts to the framework negotiated under the United Nations Framework Convention on Climate Change (UNFCCC), and how international rules inform country programs. This covers national strategies, subnational projects or demonstration activities, measurement, monitoring, reporting, and verification (MRV) systems, and finance and benefit distribution. The Chapter then evaluates different policy approaches to deforestation including regulatory, planning and market based instruments. The next section focuses on environmental and social safeguards and issues of equity and livelihoods. Linking with the discussion on safeguards, the review critiques a broad literature on rights, tenure and participation. The last section provides a critical analysis of governance research and identifies the gaps that this thesis will address. 2.1 Deforestation and GHG Emissions Deforestation, forest degradation and land use change comprise a significant source of global anthropogenic GHG emissions, second only to the burning of fossil fuels (IPCC 2013). The Intergovernmental Panel on Climate Change (IPCC) (2013, p.17) estimates annual net carbon dioxide equivalent emissions from anthropogenic land use change were 0.9 [0.1 to 1.7] GtC yr 1 (gigatonnes of carbon per year) on average during the period 2002 to 2011, equivalent to over 10% of total global emissions. Recent estimates taken from satellite data indicate that in the period approximately 2.3 million square kilometers of forests were lost annually, whilst 0.8 million square kilometers were gained, from activities such as reforestation and native regrowth (Hansen et al. 2013, p.850). These figures include loss of tropical forests from human activities and from natural causes such as fires. The regions with the highest rates include the tropical rainforests and dry forests of South America, Asia and Central Africa. Over the period Brazil recorded a significant drop in deforestation rates in the Amazon Basin (from very high levels), whilst countries including Indonesia, Malaysia, Paraguay, Bolivia, Zambia, Angola and Tanzania all recorded increases. Both Indonesia and Brazil continue to have very high rates of annual forest loss and are therefore in the top 20 countries (industrialised and developing) in terms of annual GHG emissions (UNFCCC 2014). 14

35 Estimating emissions from deforestation and other land-uses and their contribution to global warming is subject to a high degree of uncertainty. When forests are cleared and soils are intensively cultivated, carbon that is stored above and below ground in leaves, branches, stems and roots is released to the atmosphere (Baccini et al. 2012; Lorenz and Lal 2010). Carbon dioxide can also be absorbed from the atmosphere by growing forests, changing soil cultivation practices, and the regrowth of secondary vegetation after logging (LeQuere et al. 2009). Forests and land therefore function as both sources and sinks of carbon dioxide. It is often difficult, however, to separate the relatively small volumes of human emissions from the natural flux of carbon dioxide that occurs in the earth s bio-chemical cycles. These natural cycles result in the large exchange of carbon dioxide to and from the atmosphere, released and then absorbed by the world s forests and oceans. A large body of scientific research therefore aims to improve the accuracy of models and emissions estimates from anthropogenic deforestation. This research utilises data from satellite monitoring and remote sensing, national census data of forests, and other techniques in order to measure carbon biomass at different scales (Van der Werf et al. 2009; Houghton et al. 2010; Baccini et al. 2012; Hansen et al. 2013) Drivers and Causes Research conducted in different disciplines identifies a complex range of factors or drivers that contribute to deforestation. Different studies indicate various immediate causes, but the research does not outline any universal factors that cause deforestation across countries. A large proportion of this research is based on statistical modeling to correlate various socio-economic factors that drive deforestation in different countries. Whilst these studies are important, larger statistical models often have limited reference to local scale political and socio-economic issues, or other factors leading to forest loss. In terms of physical activities, agricultural expansion remains the leading contributor to land-use change and forest loss in most developing countries (Gibbs et al. 2010; Houghton 2012), whilst timber extraction, logging and fuel wood collection cause forest degradation (Hosonuma et al. 2012). Agricultural expansion involves a range of activities, such as clearing forests for intensive cropping or plantations, conversion to grazing lands and pasture, or for smallholder production. Decisions to clear forests are largely in response to underlying drivers, particularly demand from domestic or global markets, and these are often supported by proximate causes such as access to 15

36 transport (Geist and Lambin 2002; Chomitz et al. 2007). In earlier studies, poverty and population pressures were correlated with forest loss, and attributed to resource scarcity and the dependency of local communities on forest products like fuel wood (Allen and Barnes 1985). Other studies suggest issues of poverty, population growth, insecure forest tenure (Barbier and Burgess 2001a) and technology changes (Grainger et al. 2003) have some impact on forest conversion, but the effects are uncertain and not uniform in different countries or locations (Angelsen and Kaimowitz 2001). A number of studies statistically model the socio-economic factors that contribute to deforestation. Although these studies account for causes in some countries, statistical models do not adequately account for all country circumstances. Studies have focused on modeling the economic development of countries to show a correlation between low per capita GDP and high rates of deforestation, although results vary substantially between countries (Bhattarai and Hamig 2001; Barbier and Burgess 2001b; Ewers 2006). Rudel et al. (2005) developed the widely applied forest transitions model based on the environmental Kuznet s curve 2 or economic development path. Rudel et al. (2005, pp.24-25) argue that countries transition though period of high deforestation, which then slows with economic development and the correlated increase in non-farm wages. Alternatively, forest scarcity may increase the value of timber and encourage replanting. Contrasting approaches employ meta-analysis and large N comparative studies across countries to evaluate factors that drive forest conversion, such as the economic incentives available to individuals and companies (Lambin et al. 2001; Chomitz et al. 2007). There is high variation in the opportunity costs of different land uses, but forests will often be converted when there are greater returns from agricultural production. The other main drivers of deforestation or reforestation are government policies and public/private finance. Agricultural expansion in developing countries is widely supported by government policies and incentives, such as tax breaks and subsidies, settlement schemes or poverty alleviation programs (Barbier and Burgess 2001b; Chomitz et al. 2007; Pirard and Belna 2012). Likewise agricultural credit and financing arrangements encourage intensification of land use and agri-business development. Deforestation is shown to increase with government development programs, such as roads and infrastructure, sanitary programs, and agrarian reforms targeting land titles 2 The EKC (Environmental Kuznets Curve) hypothesis proposes that indicators of environmental degradation first rise with rapid economic development, stabailise and then fall with increasing per capita income. 16

37 and tenure (Pfaff et al. 2013). One of the major policy changes to occur during the 1990s was the International Monetary Fund s (IMF) structural adjustment program, and the shift from state sponsored agriculture to an enterprise model that encouraged large agri-business investment (Angelsen and Kaimowitz 2001; Rudel 2007). Structural adjustment required developing countries to liberalise trade and remove tariffs, deregulate commodity prices and exchange rates - all aimed at boosting agricultural production and exports, which in turn created further pressure on forests (Chomitz et al. 2007). In contrast to the economic modeling approaches, studies evaluate democratic quality and the relationship between deforestation rates and political conditions in developing countries. McCarthy and Tacconi (2011, p.128), for example, adopt a political economy framework and argue that there is a relationship between deforestation and the type of political regime, and how political power is exercised in the local setting. Deforestation and timber harvesting are often higher in countries with authoritarian military regimes, or where legal institutions are weak and political tenure is uncertain (Didia 1997; McCarthy and Tacconi 2011; Karsenty et al. 2013). In authoritarian countries, networks of political elites, military and other business interests often utilise forest resources and land for personal gain or to finance electoral activities. Conversely, democratisation and democratic institutions can lead to development outcomes and the introduction of sustainable forest management programs, but the results are not uniform (McCarthy and Tacconi 2011). Countries such as Brazil, Indonesia, and Canada are democracies yet continue with high rates of deforestation and timber harvesting, whereas China has a centralised authoritarian government and has recorded a net increase in forest area since the 1970s. Governance quality or capacity is also cited as a factor, and those countries with poor indicators often have problems managing forest loss and illegal logging (Umemiya et al. 2010, p.699). Corruption and collusion is also linked to resource extraction and forest conversion, often through the political lobbying process and upper level government decision making, but also in local level permit and enforcement practices (Smith et al. 2003; Barbier et al. 2005; Amacher 2006; Kolstad and Søreide 2009). Accountability deficits are particularly acute in remote frontier regions, where a lack of capacity in the forestry administration, coupled with inadequate regulations and enforcement allows illegal logging and plantation development to proceed (Tacconi et al. 2007). These governance problems often combine with other economic drivers and 17

38 remain an important focus of international and domestic efforts to address deforestation (Fosci 2012). Many of the studies on deforestation aim to identify the causal relationships between land-use change and various socio-economic, political and policy factors. This research is important in terms of understanding trends and identifying policy options, but they take deforestation emissions and the multiple causes as self evident, real and objective activities. These accounts largely ignore the processes by which different technological, scientific and social-science analysis construct the causes of deforestation. Boyd (2010, p.845) argues that efforts to manage the global carbon cycle involves the construction and deployment of new ways of seeing the problem of deforestation, often through satellite monitoring of forest cover, but also through debates about the crisis in global forest governance (see Humphreys 2006). This critique is important in terms of emphasising that deforestation can be constructed in multiple ways, and that these knowledge practices inform the different policy or institutional responses (Boyd 2010; Stephan 2012). Currently there are few detailed studies on developing countries that evaluate the complex ways deforestation and carbon emissions are framed in the national or local context, and how this informs REDD+ interventions. This thesis therefore seeks to address this critical gap by focusing on the various ways agents construct issues around deforestation, and how this shapes and limits domestic policy options. 2.2 International and National Climate Frameworks Countries negotiating the UNFCCC in 1992 recognised that deforestation and forestry management was a serious global problem. Emissions from these activities were included under the Convention and were a component of international climate negotiations over the following decades. This section provides a short overview of the development of the international framework under the Kyoto Protocol and subsequent negotiations on REDD+, before shifting to a more detailed discussion of each policy component. This discussion is important because the international climate change and REDD+ framework broadly sets out the policies and technical initiatives required from developing countries, and also the rules for emissions accounting. However, the UNFCCC framework is subject to ongoing negotiation amongst the parties, and there remains a number of unresolved issues, and a high level of discretion and ambiguity in how developing countries should implement REDD+. The detailed analysis of 18

39 Indonesia to follow in Chapters 5, 6 and 7 focuses, in part, on the application of the international framework and how this diffuses into domestic policy options. The UNFCCC recognises all sources and sinks of GHG emissions, and broadly groups a range of activities under Agriculture, Forestry and Other Land-use (AFOLU). The Kyoto Protocol to the Convention (1997) separated these activities into two sectors called Land-use, Land-use Change and Forestry (LULUCF), and Agriculture, and outlines a series of binding rules and modalities that signatories to the Protocol are required to adopt. However, a series of complex and highly contentious issues prolonged the negotiations around the LULUCF sector and resulted in a convoluted set of rules and modalities (Nobel and Scholes 2001; Hohne et al. 2007; Trines 2008; Boyd et al. 2008). As a consequence the Kyoto Protocol and related instruments provide limited incentive or opportunity for developing countries to manage emissions from deforestation. Further, the majority of developing countries are not signatories to the convention and therefore not required to meet targets under Kyoto Protocol, although a number have made voluntary commitments to undertake mitigation programs in the LULUCF sector. Under article 3.3 of the Kyoto Protocol, Annex 1 industrialised countries are required to include certain activities from the LULUCF sector in annual GHG emissions targets during the first ( ), and second ( ) commitment periods. 3 The Protocol also allows Annex 1 countries to fund forestry offset projects in non-annex 1 developing countries under the Clean Development Mechanism (CDM), which was designed primarily to provide low cost mitigation options and to fund sustainable development projects (Hohne et al. 2007). Operating rules and accounting modalities for the LULUCF sector and for implementing CDM projects were accepted at the 7 th Conference of the Parties to the Convention (CoP) held in Marrakech in 2001 (called the Marrakech Accords). The Accords currently permit afforestation and reforestation projects in developing countries, subject to detailed operating guidelines and crediting rules. Avoided deforestation projects were excluded from the CDM due to uncertainty over MRV and permanence issues, and the significant concerns raised by some Parties with the quality of forest management and governance in developing countries (Neef and Ascui 2009; Corbera et al. 2010; Lyster 2013). As of 2014, the number of forestry projects funded by the CDM is relatively small (Ecosystem Marketplace 2014), 3 Parties are required to account for emissions and removals (sequestration) from afforestation, reforestation and deforestation and can voluntarily choose to include a range of other land-use and agricultural activities (UNFCCC 2001). 19

40 with barriers cited in the literature including technical challenges, restrictive and complicated operating guidelines and high transaction costs (Sholz and Jung 2008; Anger et al. 2012) Developing REDD+ The Coalition of Rainforest Nations led by Papua New Guinea and Costa Rica first introduced a separate proposal to reduce deforestation in non-annex 1 developing countries at the Conference of the Parties (CoP) 11 held in Montreal in The program was initially called Reducing Emissions from Deforestation (RED) in developing countries and focussed on positive financial incentives and compensation to reward countries for introducing effective policies and measures. Over subsequent CoP negotiations the program expanded to Reducing Emissions from Deforestation and Forest Degradation (REDD+), incorporating a host of mitigation activities; including avoided deforestation and degradation, forest conservation, sustainable forest management, and enhancement of carbon stocks though reforestation and afforestation (Pistorius 2012; Lyster 2013). During CoP 13 held Bali in 2007 REDD+ was formerly incorporated into UNFCCC under the Bali Action Plan and subsequently elevated as a priority in the international negotiation process (Corbera et al. 2010; Pistorius 2012). The general agreement reached was that [p] arties should collectively aim to slow, halt and reverse forest cover and carbon loss, in accordance with national circumstances (UNFCCC 2007, p.17). The Bali Action Plan outlined some general principles around a phased approach for countries to design and then implement REDD+ initiatives (UNFCCC 2007). Phase 1 involves country based readiness activities including the design of national strategies, policies or action plans, and capacity building for government agencies and other stakeholders. In phase 2 countries move to the implementation of strategies and policies, and the operation of MRV systems and funding mechanisms, whilst phase 3 requires full implementation and the operation of a performance based payments system for verified emissions reductions (Gupta et al. 2013). Negotiations at CoP 17 in Cancun in 2010 and CoP 18 in Durban in 2011 agreed to a series of environmental and social safeguards focussed on issues of land tenure, forest governance, gender, and participation by stakeholders. Countries are required to address the safeguards and develop a monitoring and reporting framework. Each of these key components will be reviewed in further detail below. 20

41 The CoP 19 held in Warsaw in 2013 agreed to the Warsaw Framework for REDD+, and confirmed seven key decisions and pathways from previous negotiations. In relation to finance, the decisions reaffirm that results based finance may come from a wide variety of sources, including public and private, bilateral and multilateral. Finance can be channelled through multilateral entities such as the Green Climate Fund, and there was also agreement on incentivising non carbon benefits. The Framework outlines increased coordination and support for the implementation of activities in developing countries. Particular emphasis is placed on supporting Parties, organisations, and the private sector to take action to reduce the drivers of deforestation. Parties are to designate a national entity or focal point to serve as liaison point with the UNFCCC secretariat and other bodies. The designated national entity can also receive and obtain results based finance. In terms of MRV, national forest monitoring systems should be guided by the most recent IPCC guidelines. The UNFCCC is to develop suitable MRV modalities for anthropogenic forest-related emissions from sources and sinks, forest carbon stocks and forest-area changes. Finally, countries are to provide regular report summaries on the implementation of safeguards (UNFCCC 2013). Although the international REDD+ framework has progressed there remain a number of outstanding legal and technical issues. A major component of the UNFCCC process is resolving multiple technical details related to LULUCF emissions accounting, and the various financing options, equity and development issues (Skutch et al. 2007; Corbera et al. 2010; Olander et al. 2012). A critical element is the status of REDD+ in a future binding international climate agreement that may replace the Kyoto Protocol at the conclusion of the second commitment period in 2020 (Angelsen and McNeil 2012; Haug and Gupta 2013). Many Parties advocate REDD+ as a priority to meet shortterm international mitigation targets, and to integrate developing nations into a binding global climate agreement. However, there remains disagreement amongst the Parties in many policy areas, such as the integration of REDD+ into national climate strategies (called Nationally Appropriate Mitigation Actions - NAMAs), and the rules for trading REDD+ credits in compliance carbon markets (Olander et al. 2012). As evidenced with the LULUCF sector and the CDM, poor design and complex rules may limit the incentives that an international REDD+ framework can deliver to developing countries (McDermott 2014). 21

42 2.2.2 National and Sub-national Initiatives Although the international legal framework is incomplete, a number of developing countries have commenced national REDD+ programs, measures and activities. The majority of countries participating in REDD+ remain at the early phases focussed on preliminary policy development, technical initiatives and national strategy design (see UN-REDD 2014). Several countries including Indonesia, Vietnam, Brazil, Tanzania and the Central African Republic have progressed towards the implementation of a REDD+ institutional framework, and introduced policies to address the drivers of deforestation (see Gupta et al. 2013; Korhonen Kurki et al. 2013). The development of these national frameworks and policy measures is an important area of research on REDD+. For example, country based research has evaluated the progress, effects and results of specific REDD+ initiatives in the Democratic Republic of Congo (Aquino and Guay 2013), Cameroon (Brown et al. 2011; Somorin et al. 2014), Indonesia (Luttrell et al. 2014), and in the Amazon Basin countries of Peru (Evans et al. 2013) and Brazil (May et al. 2011; Duchelle et al. 2014). Research has also commenced on existing institutional and policy frameworks that are likely to impact or influence the design of REDD+ programs. Examples of these early critiques cover the effect of government conservation and land-use policies in Indonesia (Brockhaus et al. 2012; Lima 2013), and several countries in Latin America (Larson and Petkova 2011; Börner and Wunder 2012; Pokorney et al. 2013). In addition to national level processes, countries have commenced activities at the subnational scale. Sub-national REDD+ includes results-based demonstration activities or pilot projects that are trialling different methods and practices to conserve, sequester and store carbon in forests. These sub-national activities range from local community forestry projects, to provincial scale programs that are part of larger national initiatives. Initial studies have attempted to map REDD+ projects across countries through a large N case studies focussed on their location, design, the actors engaged, and the type of forestry activities underway (Cerbu et al. 2011; Sunderlin and Sills 2012; Lin et al. 2012). A preliminary survey by Cerbu et al. (2011) indicated that many projects are located in countries with high deforestation rates such as Indonesia, Brazil and Peru, and are often extensions to existing biodiversity and forest conservation programs. In contrast, several small N case studies evaluate the different institutional components to setting up forest carbon activities and market based projects, to determine what has proven successful in the field and to provide insights for future policies (Hajek et al. 2011; Dulal et al. 2012). Other studies focus on the capacity needs of smallholders to 22

43 successfully engage in carbon forestry and conservation programs (Caplow et al. 2011; Cerbu et al. 2013). Research on national frameworks and on sub-national projects is rapidly expanding, and provides a range of information and methodologies to evaluate various programs and projects, offering recommendations for future policy or project design. What is missing, however, is research that investigates how these multiple policies and activities operate to change and potentially transform forest and land-use administration and governance in particular countries. The critical issue is to go beyond existing evaluations and analyse how the myriad of initiatives and policy reforms actually function or operate as a system for governing resources, what they seek to achieve, and their intended effects on different forest users. 2.3 Policies to Reduce Deforestation The UNFCCC (2013) recommends countries develop suitable policies to address the drivers of deforestation and increase removals (sequestration) in forest systems. Broadly, forest policy instruments can be categorised into economic and market based policies, statutory regulations and planning instruments, voluntary forest and management programs, education and research, and technology improvements (Cubbage et al. 2007; Angelsen 2009; McDermott et al. 2010). Added to these are reforms aimed at improving public administration or the governance of forests, such as accountability and transparency in timber licensing, government capacity building, law enforcement, and the policing of illegal logging and agriculture (Kanowski et al. 2011; Amacher et al. 2012). Research on different policy mechanisms attends to the theoretical designs and applied effects of carbon market mechanisms or protected area systems, and to provide lessons learned for REDD+. Few studies, however, analyse how different policy instruments function as techniques and strategies for governing, or pay attention to why government agencies and other actors promote and select various policy instruments, and how they are employed in the context of REDD+. These gaps are explored in detail throughout this thesis Markets and Incentives Market and economic instruments comprise a range of different policy tools that are widely advocated and applied to REDD+. Recent research investigates a range of 23

44 fiscal policies that target forestry and agriculture, such as taxation systems, direct incentives and subsidies, in addition to government forest rents or revenue structures (Angelsen 2010; Angelsen and Rudel 2013; Pfaff et al. 2013). For example, governments could alter taxation systems that facilitate agricultural expansion, such as increasing levies on certain export commodities, or removing direct subsidies to small or large landholders - which are often delivered as price offsets on fertiliser or fuel (Angelsen 2010). Other policies focus on labour market conditions to encourage a transition to alternative employment, such as intensifying farm production through technology development, or supporting other sectors like tourism (Pirard and Belna 2012; Pfaff et al. 2013). More structural reforms could target economic development policies to shift focus away from timber exports, or reduce infrastructure expenditure on roads and processing facilities, which often enhance market access for agricultural products (Angelsen and Rudel 2013). However, these macro-economic policies are unlikely to be popular in countries geared towards agricultural expansion, and run counter to many development programs advanced by international organisations such as the World Bank. More targeted economic instruments include direct incentives such as payments for ecosystem services (PES). PES mechanisms are widely advocated for REDD+ at multiple scales, including local forestry projects but also large international financial transfers to developing countries (Wunder 2009; Angelsen et al. 2012; Corbera 2012; Buttoud 2012; Streck 2012; Karsenty et al. 2014). PES can be defined as a voluntary and conditional transaction between a buyer and a seller for a well-defined environmental service or land use value, requiring a legal contract or agreement (Wunder et al. 2008; Engel et al. 2008). 4 In the case of carbon sequestration, PES projects frequently involve a contract between a company and landowners to undertake specified forest conservation and management activities over a defined period, usually years or decades (Corbera et al. 2007; Wunder 2009). The theoretical premise is that payments for managing forest carbon will exceed returns from other land-use activities or logging (Engel and Palmer 2008). Despite the claims of multiple conservation and economic benefits from PES projects, the outcomes are often far from certain. For example, recent case based research indicates that PES payments can provide livelihood opportunities and services such as health care, but financial returns can be significantly lower than expected, whilst transaction and operating costs 4 PES operate according to the beneficiary pays principal, in that either public or private entities offer financial incentives to landowners or managers to protect ecosystem resources, usually for a specified time period (Wunder et al. 2008). 24

45 are often high for project developers and local communities (Corbera et al. 2011; Jindall et al. 2012; Tacconi et al. 2013; Mahanty et al. 2013). The other major economic policy instruments used for GHG mitigation are carbon markets and product information and disclosure programs. Product information schemes target existing commodity markets for timber and agricultural products to provide greater visibility for producers that meet specified standards (Pirard 2012). Examples include forest certification schemes (such as the Forest Stewardship Council - FSC), or carbon neutrality programs which require sustainable production practices from producer companies, who then market their products to ethical or conscious consumers (Pattberg 2005; Agrawal et al. 2008). Forest certification schemes are widely promoted in timber commodity chains, yet the utility and effectiveness of these policies in reducing deforestation and illegal logging remains relatively limited (Marx and Cuypers 2010). More recent programs have targeted the illegal timber trade through product verification and compliance schemes in developing countries. These include the voluntary Forest Law Enforcement, Governance and Trade (FLEGT) Agreement that utilises domestic compliance and tracking systems to certify timber products as legal. Legally sourced timber is now a requirement for access to European and North America markets (Cashore and Stone 2012; McDermott 2014). Carbon markets rely on price signals and the creation of legal rights and rules to enable trading in certified emissions credits. Under a compliance based cap and trade scheme, for example, governments set a total volume of pollution that can be emitted each year and then assign allowances equivalent to one unit of pollution (Freestone and Streck 2009; Newell and Paterson 2010). Companies and public entities are required to purchase allowances or credits in the market equal to the volume of pollution that they emit over a given period. Voluntary markets, by contrast, do not have set volumes or credit prices, but operate according to market demand for offsets (Michaelowa 2010). In both compliance and voluntary markets, credits can be generated by different activities that sequester and store carbon or reduce emissions, and this includes forestry projects or management of vegetation and soils (Pagiola et al. 2002; Portela et al. 2008; Van Kooten 2009). REDD+ projects and activities that store carbon over a defined period can therefore generate credits or units for sale in carbon markets, providing a revenue stream for the developers and contractors (Ebeling and Yasue 2008; Neef and Ascui 2009; Streck 2013). 25

46 There are multiple compliance and voluntary carbon markets in operation, including international emissions trading, and project based activities of Joint Implementation (JI) and the CDM. As of 2014 there are compliance markets operating in the European Union, New Zealand, Switzerland, South Korea and Kazakhstan, in addition to regional schemes in the United States and Canada. At the CoP negotiations held in Durban in 2011, the Parties reached the important decision that appropriate market-based approaches could be developed by the COP to support results-based actions for REDD+ (UNFCCC 2011). As of 2014, REDD+ credits were not permitted for trading in most compliance markets (Forest Trends 2014), however many market-based mechanisms are already either developed, under development, or anticipated in the future. For example, the California based cap and trade scheme is preparing options to trade in REDD+ credits. Voluntary markets do allow REDD+ offset credits, but these are not tied to compulsory emissions caps and operate as private transactions and contracts between project developers and buyers. The buyers are often large companies that participate in corporate social responsibility programs, such as energy producers or commercial airlines (Bayon et al. 2007; Michaelowa 2010; Newell and Paterson 2010) Regulations and Protected Areas In addition to carbon markets are regulatory and planning instruments. From the 1970s onwards, many developing countries adopted programs aimed at the sustainable management of public forests, however the persistent problem remains implementation, particularly in locations distant from population centres. Most developing countries have large publicly owned forest estates that are regulated though different statutory mechanisms, often to serve multiple purposes from forest harvesting and resource production, to biodiversity and watershed protection. Widely used regulatory instruments include protected area systems, resource and mining regulations, trade and export restrictions, statutory natural resource planning, licensing and law enforcement, and environmental impact assessment (Cubbage et al. 2007; McDermott et al. 2010). State controlled forests are often administered through the concession model of timber allocation, where companies pay government fees or royalties for access and rights to harvest or extract timber and other products (Gupta et al. 2013). Government agencies retain responsibility for policy setting, management and enforcement, which can be devolved or co-managed with provincial or local authorities. However, 26

47 significant deficits occur in most countries in relation to implementation of forestry regulations, due to a combination of capacity and resource constraints, endemic corruption and dysfunctional judicial systems (Tacconi et al. 2007; Kaimowitz 2008; Amacher et al. 2012). For example, timber concessions issued by government agencies often require reforestation post harvesting, but timber companies and developers routinely fail to comply with these laws and can bribe officials to avoid prosecution (Amacher et al. 2012). There is extensive literature on the design and operation of forest regulatory policies in different countries (for example McDermott et al. 2010), but currently only a cursory examination in the context of REDD+. Indeed there are multiple regulatory instruments applied to REDD+ that require further examination in particular jurisdictions, and their linkages with other components of the reform agenda. Several country studies focus on the efficacy and limitations of existing forest and land-use frameworks to accommodate REDD+, and the capacity of public administrations to manage land-use change (Brockhaus et al. 2012; Murdiyarso et al. 2012; Müller et al. 2013). Indonesia, for example, introduced regulatory measures to control deforestation, including a recent moratorium on issuing new forestry and palm oil concessions (Edwards et al. 2012; Sloan et al. 2012). Other research focuses on the optimal design and opportunity costs of different forestry and spatial planning regulations, and how they compare with alternative market or fiscal mechanisms (Angelsen 2010; Pfaff et al. 2013). A common argument cited in the literature is that improved implementation and enforcement of existing environmental regulations and land-use planning, if properly designed, should deliver deforestation benefits, as well as development and livelihood outcomes (Kaimowitz 2008; Angelsen 2009; Pfaff et al. 2013; Müller et al. 2013). However, the design of existing forestry and land-use policies may also cause problems, as overly complex legislation and laws have proven cumbersome to implement and often fail to achieve their intended purpose (MacKenzie 2013, pp ). Statutory and voluntary protected area systems are a key policy tool applied to REDD+ (Angelsen and Rudel 2013; Pfaff et al. 2014). Empirical research suggests that protected areas offer a valuable tool in reducing rates of deforestation, but design and location are important in order to target forests that are under development pressure. Most national parks and reserves are located in remote mountainous terrain, yet clearing and agricultural development occurs on arable and readily accessible land (Andam et al. 2008; Pfaff et al. 2014). Scientific research in this field utilises different 27

48 methodologies to design and locate reserves, and often with the aim of securing cobenefits such as biodiversity conservation and habitat protection (for example Phelps et al. 2012; Venter et al. 2013; Busch and Grantham 2013). Alternative policy options include integrated conservation and development programs (ICDP), and community forest management (CFM) that aim to deliver local livelihood opportunities coupled with sustainable production and/or biodiversity conservation. REDD+ projects and sub-national activities could be based on the ICDP and CFM models and provide a range of economic and social co-benefits for communities, as well as reduced rates of deforestation (Blom et al. 2010; Cronkelton et al. 2011; Sunderlin and Sills 2012; Minang and van Noordwijk 2013). Studies in Latin America, for example, argue that protected areas and CFM will be beneficial for a decentralised REDD+ approach, but need to be flexible to adapt to local conditions and be supported by robust and accountable public institutions (Oestreicher et al. 2009; Cronkelton et al. 2011; Hayes and Persha 2012; Scriven 2012). In contrast, ICDP based programs are heavily criticised for their high costs and low performance in delivering biodiversity conservation and community co-benefits, and thus may be unsuitable for the multi-level institutional structure of the REDD+ framework (Minang and van Noordwijk 2013). 2.4 Finance and Payments Finance is perhaps the most critical issue to the immediate and long term success of REDD+ at the national level but also for localised project activities. This section evaluates the different options for financing including public funding and markets, before discussing the mechanisms to distribute payments and other benefits. These issues of finance, payment mechanisms and benefit distribution are covered in-depth in Chapter 6. The original concept of REDD+ was a global scale incentive mechanism to transfer finance to developing countries in return for reducing levels of emissions against an historical baseline, and for conserving standing forests (Santilli et al. 2005; Strassberg et al. 2009). Alternative approaches (Combes Motel et al. 2009; Tacconi 2009) proposed that REDD+ could operate more as a compensation mechanism, where countries receive finance for implementing a set of agreed policies, and to offset opportunity costs from foregone timber and agricultural production. The following statement summarises the REDD+ incentives approach: 28

49 REDD+ is an incentive-based instrument, based on the idea that both public and private agents are self-interested and are able to calculate the full cost and benefits associated with various options. The basic idea is that developing countries have an opportunity cost if they choose to conserve (in the broad sense) their forests rather than convert them to agriculture or any other nonforest land use. The REDD+ mechanism is intended to provide sufficient financial incentive to change those public or private decisions that would otherwise lead to forest conversion. This means that the amount of transfers either through the carbon market or through international ad hoc funding is comparable with the opportunity cost, which differs widely among countries (Karsenty et al. 2014, p.22). REDD+ thus requires a large and dedicated financing stream that will meet the needs of developing countries (Streck 2012; 2013). Parties to the UNFCCC negotiations have agreed to provide additional and longer term finance for REDD+, and that all options should be available including public funds, market finance and private sector investment (UNFCCC 2012). Annex 1 developed countries have made significant pledges to provide REDD+ funding, whilst developing countries have committed to financing some forestry programs from domestic budgets. In addition to these traditional public fund options, carbon markets are widely advocated as the major source of longer term finance, particularly if they attract private investment (Elliasch 2008; Neef and Ascui 2009; Corbera et al. 2010; Karsenty et al. 2012; Streck 2012). Compliance markets tied to international and national emissions caps could provide tens of billions of dollars annually, either as direct payments to project activities or as large fiscal transfers to national governments (Streck 2012). The current major sources of REDD+ start-up finance are from multi-lateral climate funds and from official development assistance (ODA), with a small percentage from voluntary carbon markets. For example, the Government of Norway provides substantial funds for REDD+ activities in Indonesia, Brazil, Guyana, Mexico and Peru (Streck 2013). Other major REDD+ donors include Australia, Germany, UK, Japan, USA and the European Union, who channel much of their funding through existing ODA programmes, or multilateral climate funds. Multilateral sources include the United Nations Collaborative Programme on Reducing Emissions from Deforestation and Forest Degradation in Developing Countries (UN-REDD Programme), the World Bank Forest Carbon Partnership Facility (FCPF) and Forest Investment Programme (FIP) 29

50 5,the Global Environment Facility (GEF), and the newly established Green Climate Fund under the UNFCCC. In 2012, these funds had collectively pledged upwards of US $1.5 billion (Streck 2012, p.659). Multilateral funds currently allocate finance to national governments for agreed activities and measures, and these payments are in turn re-allocated to sub-national programs and initiatives. Countries have used this start-up finance for readiness inputs such as capacity building, policy and technical reforms. Once REDD+ policy frameworks are in place, and payment and MRV systems are operational, countries are required to move to a performance based financing model Benefit Distribution Finance requires institutions to disburse payments, including the mechanism and rules, and an authority accountable for managing the funds. As part of phase 1 readiness activities countries are required to develop national, sub-national and project based payment mechanisms. There are several options available including existing multilateral climate funds and ODA processes, domestic grants and government budgets, as well as contracts under a market or PES type scheme (Vatn and Angelsen 2009; Hoang et al. 2013; Pham et al. 2013). Payment or benefit distribution mechanisms are therefore a key component of the REDD+ institutional architecture at the national and sub-national level. Hoang et al. (2013, p.49) define a benefit distribution system as a set of rules, regulations, policies and mechanisms for sharing monetary and non-monetary rewards from policy or market incentives. Benefit distribution systems determine how incentives or rents are delivered to government agencies or individuals, the range of programs or projects funded, and the application of specific co-benefits and safeguards (Karsenty et al. 2012; Vatn and Vedeld 2013). REDD+ benefit and payment distribution can incorporate a vertical axis operating between jurisdictional levels, but also a horizontal axis between stakeholders at the national level, or locally amongst communities and households (Pham et al. 2013, p.1). 5 The UN-REDD Programme was launched in 2008 between the Food and Agriculture Organization of the United Nations (FAO), the United Nations Development Programme (UNDP) and the United Nations Environment Programme (UNEP). The Programme supports national REDD+ readiness efforts in 48 countries in Africa, Asia-Pacific and Latin America, in two ways: (i) direct support to the design and implementation of UN-REDD National Programmes; and (ii) complementary support to national REDD+ action through common approaches, analyses, methodologies, tools, data and best practices. The FCPF was launched by the World Bank in The FCPF readiness program is currently supporting 37 developing countries to build REDD+ strategies, MRV systems, and national baselines. The FCPF also operates a Carbon Fund to provide performance-based payments for verified emission reductions. FIP is a component World Bank World Bank Climate Investment Fund, operational since July Designed to provide scaled-up financing for forest sector reforms identified through national REDD+ strategies. 30

51 There is substantial debate on the design, efficiency and effectiveness of various distribution mechanisms (Obersteiner et al. 2009; Neef and Ascui 2009; Strassberg et al. 2009). Vatn and Vedeld (2013, p.429) argue that a national public fund provides the most effective option to target deforestation, yet critical issues remain with government transparency and accountability. Other issues relate to the efficient allocation of public funding. Many countries have experienced problems with multilateral and bilateral funds that have proven cumbersome to operate and slow to disperse REDD+ payments (Streck 2013). There is also debate over the governing body or fund manager, and whether the national government should have ultimate authority, or operate in a formal partnership with multilateral agencies (Vatn and Angelsen 2009; Pham et al. 2013). Carbon markets and PES programs may offer a more efficient mechanism to allocate incentives, with contracts providing a direct process to exchange payments between buyers and sellers of forest offsets (Wunder et al. 2008; Peskett et al. 2011). But carbon markets operate with large uncertainties, such as credit demand and price volatility, and there are liabilities on governments or project stakeholders in terms of meeting contract conditions and managing stored carbon over extended time intervals (Isenberg and Potvin 2010; West 2010; Corbera et al. 2011; Palmer 2011). Benefit sharing is central to the incentives approach that underpins REDD+ and will determine and type and extent of interventions undertaken. Distribution mechanisms also have a complex range of performance and normative objectives, such as the equitable delivery of incentives and benefits to stakeholders, improved policy and administrative outcomes, and requirements for fiscal accountability (Pham et al. 2013). Rules for benefit sharing have important implications in relation to tenure and rights, the provision of services and community development, and fair compensation for affected parties. At the national level, benefit distribution is dependent on existing legal and administrative frameworks and budget processes for allocating public funding, and is often integrated with other policy priorities around economic development or service delivery. There is a body of research that evaluates the design of distribution options in relation to PES and carbon forestry projects, identifying options or lessons for REDD+ (for example Wunder and Alban 2008; Peskett et al. 2011; Tacconi et al. 2011; Karsenty et al. 2014). Preliminary research has looked at existing mechanisms, such as the CDM to inform national REDD+ development (see Lederer 2011; Costenbader 2011), but there are few studies that critically evaluate the role of benefit distribution in REDD+ regimes in specific countries. Apart from cross country research by Pham et 31

52 al. (2013) and by Hoang et al. (2013) on Vietnam there is limited research that assesses the design of national and sub-national REDD+ benefit distribution mechanisms, and how they operate as an instrument of government intervention. Research is therefore required on the relationship between benefit distribution and the range of other components to the REDD+ agenda, and with the broader political objectives around economic and social development, rights and equity. These issues will be addressed further in Chapter Carbon Accounting The allocation of finance and benefits requires suitable and robust carbon accounting systems, a key requirement of international agreements on REDD+. Carbon accounting and measurement issues thus remain at the forefront of international negotiations and national/sub-national programs. The LULUCF sector and now REDD+ are subject to a number of complex and often controversial measurement and accounting problems. These issues require resolution as part of finalising an international climate agreement, and before countries are eligible to receive performance based payments, but also to allow trading of REDD+ credits in compliance carbon markets (Corbera et al. 2010; Atmadja and Verchot 2012). The three key issues subject to wide discussion in the literature are non-permanence, leakage and additionality (Korhonen-Kurki et al. 2013, p.345). Non-permanence relates to the risk that emissions sequestered in forestry projects can be easily reversed after a change in land use, harvesting or damage from fire and disease (Schlamadinger et al. 2007; Devries and Herold 2013). 6 Leakage could occur if, for example, a carbon offset project established in one location shifts deforestation and land use activities to another area, resulting in no net reductions or an overall increase in emissions at the national level (Scholtz and Jung 2008; Atmadja and Verchot 2012). Projects will also need to prove additionality, such that an activity or measure has reduced emissions above and beyond what would have occurred without the intervention, and is therefore additional to pre-existing policies or activities (Devries and Herold 2013, p.162). Other accounting issues relate to how countries should measure and report on national REDD+ activities, and if projects should be part of these national accounts in a type of nested approach (Pedroni et al. 2009). 6 Forests and soils also reach a level of saturation over time, and projects will require long term management to maintain carbon stocks, which may prove to be untenable in some developing countries. 32

53 2.5.1 MRV Systems According to UNFCCC guidelines, countries are required to develop REDD+ MRV systems and reporting mechanisms, often as part of national GHG accounting frameworks that cover all emissions sources and sinks 7. For REDD+, the UNFCCC requires robust and increasingly sophisticated MRV systems that utilise accounting methodologies and standards designed by the IPCC. The purpose of MRV systems is to accurately track and monitor all changes in the carbon stocks of forests caused by human activities. In addition, countries are required to prepare a REL, or a baseline estimate of deforestation emissions from which to measure the effectiveness of any interventions (Devries and Herold 2013, p.157). Policy implementation increasingly relies on a national reference emissions level (REL) and MRV systems, which underpin the transfer of funding and payments to countries (and to specific activities), but also to regulate credit transactions in carbon markets (Herold and Skutch 2011; Murdiyarso et al. 2012). Monitoring is therefore essential in order for countries to be compensated for any emissions reductions. MRV systems are also required for assessing safeguards (discussed below), policy implementation, and co-benefits such as economic development and biodiversity conservation. MRV thus functions as a broader instrument to track carbon in the landscape in addition to assessing a range of policy, governance, social and environmental indicators. Measuring and monitoring deforestation and forest degradation is a complex undertaking and there are numerous limitations with current methods. Detailed scientific research therefore focuses on the design of carbon MRV and REL methodologies, often coupled with options for setting-up appropriate national systems and institutions. Technical approaches have focussed on the methods to collect and analyse forest data in developing countries, such as through satellite monitoring, often combined with remote sensing techniques to measure changes in forest cover (Baker et al. 2010; Asner et al. 2013; Devries and Herold 2013). Other studies focus on the methodological and organisational design for MRV systems and RELs, often providing recommendations for countries with limited forest monitoring capabilities (Herold and Skutch 2011; Bucki et al. 2012). Evaluations indicate there are technical and capacity shortfalls in most developing countries, in terms of data collection and management, 7 The IPCC proposed a tiered approach for MRV: Tier 1, the simplest to use, is based on globally available activity data and default emissions values, Tier 2 is based on country - or region-specific data, and Tier 3 uses higher-order, spatially explicit data to obtain greater certainty. Under UNFCCC rules developing countries are requested to develop the capacity for Tier 3 monitoring (UNFCCC 2011b). 33

54 and in calculating GHG emissions at standards that meet IPCC guidelines (Romijn et al. 2012; Murdiyarso et al. 2012). Administrative and skills shortages, coupled with the high costs of establishing sophisticated forest monitoring programs, means most developing countries will struggle to have operational MRV systems in place. Project developers are also designing and testing localised carbon MRV and REL. These serve several purposes, primarily to support the operation of the project, but also to provide data and information to feed into national MRV systems. Cross-case evaluations in different countries suggest that capacity varies across projects and countries, but many project proponents have made significant progress towards robust MRV systems (Shijo et al. 2013). A range of authors have argued that project based MRV should provide employment for local communities, as most data collecting is currently carried out by forestry professionals (Palmer-Fry 2011; Larraźabal et al. 2012; Pratihast et al. 2013). Potentially there are multiple benefits from this community REDD+ monitoring, such as the provision of training for local households, expanding access to forest areas, and encouraging active engagement in government interventions (Larraźabal et al. 2012). Further, community based monitoring can reduce operating costs and provide accurate, reliable and efficient on-ground data collection in remote forest regions (Pratihast et al. 2013). However, one large multicase analysis shows that most REDD+ projects for the voluntary market engage few communities in MRV, suggesting a substantial gulf between international commitments and local implementation (Danielsen et al. 2013). These carbon accounting and MRV systems are among the most critical elements of any national program. Governments and other agents will potentially use MRV systems to design, prioritise and assess a range of REDD+ interventions. Current research on carbon accounting and MRV tends to focus on the scientific and technical capacity issues of establishing and operating MRV systems. Contrasting critiques focus on carbon accountability as type of scientific knowledge system that constitutes forests as an object of government activities. These studies argue that MRV systems often format complex forest systems into simplified and comparable carbon spaces that are rendered amenable to political actions (Gupta et al. 2012; Lövbrand and Stripple 2012; Lovell 2013). In the context of REDD+, further research is required to understand how MRV systems operate within countries as a knowledge and technical apparatus that determines how areas of forest, land and related populations can be subject to certain political, regulatory and economic interventions. These issues and gaps will be addressed in detail in Chapters 5 and 8. 34

55 2.6 Social and Environmental Safeguards Carbon accounting is tied to broader issues and risks regarding the viability of REDD+ actions. Projects and wider policy reforms have the potential to deliver perverse outcomes, and pose substantial risks in some countries, particularly to local communities and forest ecosystems. The concern with potential negative effects from REDD+ has led to the development of international safeguards. These REDD+ safeguards refer to a collection of standards and rules relating to social co-benefits, democratic and governance capacity, and minimising adverse environmental impacts. Participating countries are required to implement safeguards, develop a monitoring framework, and report annually on their status. A key concern raised by different countries and interest groups in UNFCCC negotiations is whether REDD+ policies and projects can deliver co-benefits, such as sustainable development and human rights (Visseren Hamakers et al. 2012). Stakeholders have also identified potential risks or tradeoffs in REDD+, such as loss of economic opportunities and increased poverty, as well as perverse outcomes from poor program implementation (Huettner 2012; Jagger et al. 2012). In response, Parties to the UNFCCC (2011) developed broad guidelines around social and environmental safeguards outlined in the Cancun Agreements; however rules and protocols on implementation remain at the discretion of participating governments. The UNFCCC REDD+ safeguards include; the adoption of policies that are consistent with national legislation and forestry programs; the implementation of effective forest governance structures; and policies that enhance biodiversity conservation and prevent displacement of emissions (UNFCCC 2011; Lyster 2013). Additional social safeguards focus on gender equity considerations, land tenure issues, and the full participation of relevant stakeholders, with an emphasis on indigenous peoples and local communities. Environmental safeguards aim to ensure that activities are consistent with conservation goals and to prevent perverse outcomes (UNFCCC 2010). Scientific and policy research on environmental co-benefits and safeguards focuses on the potential synergies between carbon management and biodiversity protection. Global modelling studies suggest that slowing the rate of deforestation though REDD+ could deliver other ecosystem services like watershed protection and prevent further species extinctions (Strassberg et al. 2012). Other studies, by contrast, warn that REDD+ projects may not always deliver other environmental benefits, particularly if the focus of incentives is carbon content instead of biodiversity (Phelps et al. 2012, p.540). Studies 35

56 on environmental safeguards emphasise the need for policy design at the national level to address potential trade-offs, arguing that different strategies could focus on either combined or separate carbon/biodiversity outcomes (Harvey et al. 2010; Phelps et al. 2012; Potts et al. 2013). For example, Gardner et al. (2012) argue for spatial mapping techniques to integrate biodiversity priorities into land-use planning, whilst Harvey et al. (2010) call for the bundling of carbon payments with biodiversity incentives. The debate over social safeguards is integrated closely with the rights of forest dependent communities and indigenous peoples, their livelihoods and in many cases cultural survival. Lyster (2011, p.120) argues that the social safeguards agreed at Cancun were ground breaking by recognising forest people s rights and making explicit reference to the United Nations Declaration on the Rights of Indigenous People. 8 The rights agenda covers a broad collection of legal and normative issues around access and ownership to land, economic opportunities, recognising local cultural and social practices, and importantly the responsibility of countries to implement these rights (Sikor et al. 2010; Ribot and Larson 2012). Savaresi (2012, p.9) contends that the introduction of safeguards addresses some of these human rights, but there is a need further develop clear rules and procedures for their application across multiple countries. Critically, the recognition and implementation of these rights remains a domestic responsibility, leading to objections that REDD+ could further alienate indigenous people in countries where governments routinely fail to recognise, and at times actively suppress, local interests and customary practices (Lematre 2011). The introduction of social safeguards has encouraged research on their formation, the actors and networks involved, and their implications in terms of delivering equity and benefits for local and/or indigenous communities (Chhatre et al. 2012; McDermott et al. 2012; Visseren Hamakers et al. 2012; Krause and Nielsen 2014). For example, McDermott et al. (2012, p.65) develop a typology to assess how government agencies and non-state organisations have designed safeguards and incorporated them into REDD+ operations. They compare how major financiers such as governments and international organisations prioritise safeguards for carbon markets, accounting and governance reforms, whilst networks of non-government organisations (NGOs) have 8 United Nations Declaration on the Rights of Indigenous People states explicitly that: people cannot be forcibly removed from land; should be consulted about policies and legislation that may affect them; are able to make decisions, plan and develop resources; and maintain rights to access and secure subsistence from land owned or occupied. Most developing countries are signatories to other international conventions on human rights that translate to REDD+ including; the Convention on the International Covenant on Civil and Political Rights (ICCPR); and the International Covenant on Economic, Social and Cultural Rights (ICESCR). 36

57 focussed more on indigenous rights and social justice issues (McDermott et al. 2012). They argue that international safeguards operate successfully as a boundary object by enabling disparate actors to progress various proposals and ideas, and then have these translated into operational rules or programs of REDD+. In contrast, Chhatre et al. (2012, p.658) argue that safeguards have progressed from a safety-net approach into a normative program that helps to build social co-benefits at the community level. The authors contend (Chhatre et al. 2012, p.654) that safeguards can be empowering for forest dwelling communities who utilise the framework to gain access rights to forests, and increase participation in decision making Stakeholder Participation and Engagement Multi-stakeholder engagement is a cornerstone to the introduction of national and subnational programs. Participation represents a key safeguard under international REDD+ agreements, and more specifically the full and effective engagement of indigenous peoples (UNFCCC 2011). International rules also require countries to develop and implement free prior and informed consent (FPIC) as a condition of any interventions that affect indigenous communities. The majority of multilateral and bilateral REDD+ programs, as well as NGO supported local projects, require multistakeholder engagement and significant resources are committed to facilitating these activities. Participation extends beyond the local context to incorporate multiple public and private stakeholders in a complicated array of processes, covering indigenous groups in the international negotiations (Schroeder 2010), national policy reforms and strategy formation (Thomson et al. 2011), local participation in market activities and projects (Corbera and Brown. 2010; Awono et al. 2014; Cromberg et al. 2014), MRV systems (Danielsen et al. 2013), and in safeguard design (Krause et al. 2013). Similar to other areas in this field, there is a growing body of research on participatory processes in REDD+, with a strong emphasis on procedural quality and whether participation delivers benefits and equity for communities. Participation and engagement focus on the application of democratic norms that are widely considered central to an effective, equitable, efficient and legitimate REDD+ process (Lawlor et al. 2010; Thomson et al. 2011; Chhatre et al. 2012). Stakeholder engagement supports the democratic rights of local communities in REDD+, but also functions to improve transparency and accountability in forest related governance (Lawler et al. 2010, p.5). Research in this field has focussed on the procedural quality and effectiveness of participation, often framed as input and output legitimacy (Krause 37

58 and Nielsen 2014) (see further discussion below). For example, Lawler et al. (2010, pp.5-6) argue that in order to address institutional requirements, participatory processes should be part of tenure resolution processes, land titling, planning and other activities that affect community livelihoods. Stakeholder participation is also important to the output legitimacy, or the success and effectiveness of REDD+ strategies and projects. Engagement of local communities is likely to improve understanding and acceptance of REDD+ interventions, and reduce project transaction and operating costs by providing a localised source of labour for carbon monitoring and activities (Chhatre et al. 2012; Larrazábal et al. 2012). Recent empirical research on REDD+ projects aims to evaluate whether developers are supporting the goals of participation. This research applies principles of equity and distributive justice, arguing that participatory processes are important for delivery of cobenefits. These studies emphasise the value of establishing or improving institutions so that they are representative and inclusive of multiple interests (Corbera and Brown 2003; Corbera et al. 2007; Lawler et al. 2010). For example Lawler et al. (2013) develop a benefit and participation framework to evaluate five projects that apply FPIC. Results indicate that FPIC processes and consultation are highly variable across projects, and although stakeholders are aware of REDD+, they often remain ill informed of project details or the conditions of carbon contracts (Lawler et al. 2013, p.300). Krause et al. (2013) found similar results with participation amongst indigenous groups in the Ecuadorian Amazon, where communities were generally aware of the REDD+ conservation program, but had limited understanding of contracts and agreements. Awono et al. (2014) interviewed local communities about tenure resolution processes in two REDD+ projects in Cameroon. Results suggest that participation varied and produced mixed results, with around half of the villagers engaged throughout the project phases, whilst others were only involved in the latter stages of implementation (Awono et al. 2014, p.83). Other studies consider the quality of participation, with criticism directed at the institutions and the application of different processes by state and non-state organisations (Luyet et al. 2012; Korhonen-Kurki et al. 2012). Marion Suiseeya and Caplow (2013) conducted a multi-case study of 56 REDD+ projects verified under the Climate Community and Biodiversity Alliance, a third party standard that verifies and accredits projects for the voluntary carbon market. Applying a procedural justice framework the authors (Marion Suiseeya and Caplow 2013, p.975) find that although the standards require extensive consultation and participation, in the majority of 38

59 projects the proponents fail to adequately involve stakeholders in decision making. In terms of national level processes, Thomson et al. (2011, pp ) argue that current engagement strategies are limited and top-down, and often include only those stakeholders familiar with international initiatives and exclude more marginal groups. Thomson et al. (2011, pp ) suggest the only way that national programs and measures will successfully mitigate emissions is by broadening participation in order to align multiple interests with the goals and activities of REDD+. The conclusion to draw from these studies is stakeholder participation remains an important goal of REDD+, but the execution is often problematic and shaped by broader socio-political processes, results are context specific, and the provision of co-benefits is often marginal or difficult to ascertain. Current research on participation is valuable and provides a number of insights on the processes employed in different countries and locations. Critically, however, the current focus on rights provides limited insights as to why state and non-state agents deploy multiple participatory processes as a strategy to govern people and forests. Understanding the procedural and output qualities or legitimacy is important, but this needs to be evaluated in the context of what participatory mechanisms seek to achieve, how they function as part of a suite of REDD+ initiatives, and more fundamentally what is the rationale behind their application. This is particularly important in developing countries where stakeholders struggle to have input in the policy process, and are periodically subject to exclusion by governments and private interests. Participation and engagement strategies are therefore the focus of analysis in Chapter Equity, Livelihoods and Forest Tenure Participation is also linked to issues around equity, another cross cutting theme in research on forest carbon projects and REDD+. Equity relates to a broad suite of issues and related discourses about the effects of institutional processes on rural communities, and who benefits from carbon initiatives (Luttrell et al. 2012). There are several different approaches or frameworks used to evaluate equity in forest carbon initiatives. Corbera et al. (2007, p.368), for example, propose a three tiered approach to evaluate the equity of PES based carbon projects and focus on access to resources, decision making authority and outcomes in terms of benefits received. The authors (Corbera et al. 2007, p.378) conclude that the actual design of PES programs is 39

60 important in terms of household benefits, and that projects need to incorporate nonformal or customary institutions in order to deliver equity and avoid entrenching existing inequalities. In addition, McDermott et al. (2013) provide an equity evaluation framework based on principles of justice that incorporates distributive, procedural and context based components. They similarly conclude that carbon market interventions designed solely to distribute payments without attention to local inputs, access to resources and community relations, are unlikely to deliver equitable outcomes (McDermott et al. 2013, p.424). The debate over equity also relates to livelihood benefits of forest dependent communities. Livelihood is a recurrent theme in forestry studies, as the rural poor often remain excluded and marginalised by state sanctioned timber and land development policies, as well as poorly designed conservation programs (Tacconi et al. 2013). REDD+ has the potential to further exacerbate poverty if interventions lead to further exclusion from forests or prevent alternative livelihood and development programs. More broadly, the current global governance structures proposed for REDD+, and particularly the emphasis on markets and monetary payments, may fail to deliver localised benefits, and instead increase commodification of forests and erode the rights of communities (Gupta 2012, p.624). Ribot and Larson (2012, p.247) argue that governments may therefore need to introduce policy changes to address imbalances, and shift markets and institutions to be responsive to community needs and demands. A number of recent cross case studies in Africa have raised important questions regarding project objectives and implementation, arguing that pilot REDD+ interventions often fall short of providing adequate income incentives or other social benefits (Groom and Palmer 2012; Mutalahati et al. 2012). Case studies from countries in Latin America and Asia expose how new carbon based PES institutions can create distributional problems and entrench inequalities, with some politically connected families able to control critical information and access to rewards, such as employment and training (for example Kosoy et al. 2008; Chhatre and Agrawal 2009; Milne and Adams 2012; Pokorny et al. 2012). Studies in the Amazon Basin, for example, indicate that current REDD+ policies often repeat past mistakes by failing to align environment programs with community development needs, or local forest production and management systems (Pokorney et al. 2013). Pokorny et al. (2012) offer a more scathing critique arguing that market based carbon projects have largely failed to fulfil development expectations, and the attractiveness to communities of the proposed forest management activities were limited. The authors contend that 40

61 [c]ritically, it became obvious that attempts to ensure smallholders competitiveness in markets implied the gradual replacement of locally adapted production schemes, and tended to create new relations of dependency reinforcing the traditionally existing paternalistic structures (Pokorny et al. 2012, p.397). These studies support the theme that carbon based projects often fall short of responding to issues of equity, or meeting the needs of the rural poor Tenure and Carbon Rights Tenure and rights over forest carbon are at the forefront of debates on the equity and legitimacy of REDD+. Tenure has important implications for the effectiveness and efficiency of projects at the local scale, and whether they can successfully deliver carbon mitigation (Sunderlin et al. 2009; Angelsen et al. 2012). Complex laws and land allocation systems in many countries mean that forest tenure is uncertain and highly contested, which often leads to dispossession of local indigenous communities. Processes to resolve tenure and carbon rights in the context of REDD+ therefore have the potential to impact directly on communities and households. Similar to other areas, tenure and carbon rights have received substantial attention from researchers, focussing on existing legal frameworks and the policy changes required to make REDD+ equitable and effective. However, tenure is also an important instrument for governing land, as it determines where carbon projects can occur, and how the state, private and communal interests claim or demarcate access to resources. Research therefore needs to place tenure issues and processes within the overall framework of forest policy and governance, benefit distribution, and the shift to markets in ecosystem services. These issues of tenure and marketable carbon rights are analysed further in Chapter 6. The concept of land and forest tenure is complicated and disputed, but has been defined as the right, that determines who can hold and use land (including forests and other landscapes) and resources, for how long, and under what conditions (Doherty and Schroeder 2011, p.71). Most countries have a tenure system that establishes the legal rights to access, use and trade in forest resources. These statutory tenure arrangements are based on individual property rights and codified in state law, where land is recognised as an asset that can be bought and sold in a market (Doherty and Schroeder 2011, p.72). Also called de-jure rights they are determined and protected by the state, through such mechanisms as registered land titles, forest concessions and contracts. However, rights to carbon in forests are a relatively new concept and 41

62 many countries do not have adequate laws in place. The legal relationship between statutory land tenure and rights to forest carbon are important for operating and marketing REDD+ projects. Creating a tradeable ecosystem commodity like carbon often requires legislative reforms that establish and allocate property rights, and define the principles of ownership for any emissions credits (Palmer 2011). Importantly, these laws need to assign liability to either the state or project operators in the event of carbon reversal from any reduction in a forest area (Palmer 2011, p.574). Many forest dwelling communities claim customary tenure or non-dejure rights that are often outside formal legal statutes. Customary tenure is characterised by norms and oral agreements over the utilisation and ownership of land, including communal and collective use of forests (Larson 2011). In many tropical developing countries, a large proportion of forest areas are legally declared as public assets under state administration, yet access is often governed by informal customary systems. This leads to confusion and often conflicting claims over forested land. Several multicountry studies (Larson 2011; Larson et al. 2013; Sunderlin et al. 2014) evaluate existing tenure systems and their implications for REDD+. Some countries have taken steps to legally recognise customary tenure and large areas of forest have been transferred to community stakeholders. However, there are often insufficient institutional arrangements to support and enforce the rights, and governments regularly attempt to limit or reverse the reforms (Larson 2010). With the exception of Brazil, most countries lack the appropriate legal mechanisms or processes to address communal rights to forests under REDD+, or resolve competing claims from commercial interests, other villagers or the state (Larson et al. 2013; Sunderlin et al. 2014). Sunderlin et al. (2014, pp.47-48) conclude that most countries need to improve the legal mechanisms to clarify tenure, provide processes for consultation and conflict resolution, and integrate national and local processes. REDD+ projects need to address customary tenure at the local level, as it relates directly to other outputs such as current and future land-uses, livelihoods, and commercialisation of carbon. Many projects are attempting to resolve tenure in a climate of competing land-use pressures. This is compounded by a lack of legal clarity and overlapping titles or claims, corrupt practices and intimidation, ill-defined forest boundaries, and out-dated or non-existent mapping (Naughton-Treves and Wendland 2014; Sunderlin et al. 2014). Recent small N case studies on REDD+ projects by Duchelle et al. (2014) in the Brazilian Amazon, Resosudarmo et al. (2014) on Indonesia, and Awono et al. (2014) in Cameroon reveal a wide diversity of land-use 42

63 and tenure claims. Strategies and processes that successfully address tenure issues often involve extensive negotiation and collaborative agreements between local communities, project proponents and government agencies. These studies also point out that cash payments made directly to communities in return for managing and restoring forest areas are unlikely to work in areas that have insecure tenure and remain open to incursions (Duchelle et al. 2014; Resosudarmo et al. 2014). There is a level of consensus that forest carbon projects create significant challenges for communities that hold only customary rights (Corbera et al. 2011; Larson et al. 2012; Naughton-Treves and Wendland 2014). In the absence of statutory legal rights over forests, local communities may be unable to claim the carbon on their land, leading to disempowerment, loss of income and restrictions on access to forests (Dressler et al. 2012, p.689). Conversely there is also some optimism that carbon projects could benefit communities by clarifying and securing forest tenure, strengthening local identity and empowering communities. Research has also shown that obtaining legal tenure leads to a sense of ownership, and communities then have a vested interest in preserving and sustainably managing the forest resource (Chhatre et al. 2012). Securing tenure and increasing rights to forests and carbon could therefore reduce local deforestation rates, however results in different countries remain inconclusive (see Naughton-Treves and Wendland 2014). Barbier and Tesfaw (2012, p.891) support the argument that if REDD+ helps secure communal and customary tenure it could address poverty and improve livelihoods, and is thus critical to the success or output legitimacy of projects. The consensus is that tenure reforms are critical to a whole raft of REDD+ outcomes including the development of carbon commodities, and the capacity of project managers to enforce carbon contracts. Importantly, tenure will also influence the impact of these projects on socioeconomic outcomes and the recognition of cultural and democratic rights. 2.8 Governance Approaches to REDD+ The preceding sections reviewed research on a number of the specific elements to REDD+, from international frameworks to national policy and project activities. Many of these initiatives are analysed through various theoretical approaches to governance. The purpose of this section is to critically review the application of governance theories in the REDD+ literature, and to identify the limitations and gaps with these approaches. It is beyond the scope of this review to cover the diverse body of material on climate 43

64 change governance at the international and increasingly regional, national, and subnational scales. Several common theoretical approaches emerge in the climate change and REDD+ governance literature. Notably a focus on institutional architecture and the interplay between levels and activities; the agency of state and non-state actors; and the relationship between agency, discourse and institutional design. In addition, research has focussed on governance as a normative program designed to improve forestry administration and management in developing countries. Each of these approaches will be covered below Climate Change Governance There are multiple theoretical orientations to the analysis of climate change governance. On a global scale, these studies draw on theories and concepts from the broader governance literature. According to Rosenau (2000, p.172) global governance is exercised through the coordinating activities of states, but also through a vast array of rule systems, that exercise authority in the pursuit of goals and that function outside national jurisdictions. Governance also focuses on the plurality of governing agents, but importantly that non-state political actors have expanded their agency outside formal institutions of the state and intergovernmental regimes (Biermann and Pattberg 2008). Governance theories therefore pay attention to the particular locus of authority, the specific actor configurations or networks, and the legitimacy of certain governance arrangements. The term governance is also a process that emphasises new modes of steering of society and markets in contrast to the rowing conducted by a centralized state administration (Pierre and Peters 2000). Governance focuses on the interdependencies between political authorities and social and economic actors capturing the policy networks, public-private partnerships and informal decision making processes that emerge out of the interactions between bureaucracies and civil society. In the climate change context, governance approaches often focus on the processes and institutional arrangements that state and non-state agents use to steer or guide multiple initiatives or actions. A commonly cited definition of climate governance is all purposeful mechanisms and measures aimed at steering social systems towards preventing, mitigating or adapting to risks posed by climate change (Jagers and Stripple p.388). This definition covers numerous mitigation and adaptation activities that are significant both in establishing international rules and institutions, and 44

65 shaping policy at the domestic and local scales. Climate governance can therefore incorporate any number of programs operating at various scales such as international agreements, to mandatory and voluntary carbon markets, transnational climate programs, regional partnerships, domestic laws and programs, and the provision of climate science and information (Bulkeley and Newell 2010). The international climate regime operating under the UNFCCC and the interactions between nation states provides one type of governance arrangement (Young et al. 2008; Bulkeley and Newell 2010). Studies in regime governance have sought to understand the interplay between the multiple levels of political authority, and the interactions between the various international institutions and the territorially bounded entities operating under the nation state (for example Gupta 2007). Contrasting theoretical applications have sought to evaluate an emerging series of adaptation and mitigation initiatives that operate within and beyond the confines of international, state, and sub-national frameworks (Bulkeley and Newell 2010). For example, global to local networks of cities for climate change protection functions as a formal governance process that is outside the interstate regime (Kern and Bulkeley 2009; Bulkeley et al. 2014). Other approaches examine transnational climate governance involving networks of actors that coalesce around different institutional arrangements within and outside the state (Pattberg and Stripple 2008; Andanova et al. 2009; Bernstein et al. 2010; Bulkeley et al. 2012). A key feature in transnational climate governance is the formation of new types of agency operating within quasi-private institutional domains; such as the corporate governance of carbon neutrality programs or the design and operation of standards for the voluntary carbon market (Pattberg and Stripple 2008; Bulkeley et al. 2012). The multiple climate and carbon initiatives are also conceptualised as different modes of governance. Modes relate to institutional structure or design and which state and non-state actors exercise authority in the complex array of processes and activities (Bumpus and Liverman 2008; Pattberg and Stripple 2008; Bäckstrand et al. 2010; Bulkeley 2012). Different modes of governance are also associated with the emergence of new types of policy instruments, such as economic incentives and nonlegally binding programs, in contrast to traditional government regulations and legal sanctions (Jordan et al. 2003; Treib et al. 2007). In climate change policy, older modes of governance refer to public programs and technical initiatives conducted through the hierarchical administration of the UNFCCC or individual states. Newer modes refer to carbon markets, or multi-actor networks that operate through more horizontal and 45

66 flexible arrangements, where authority and climate initiatives rest on various private or civil society actors (Bumpus and Liverman 2008; Newel and Paterson 2010). However, governance modes are often complex. For example, market instruments such as the CDM function as a type of hybrid multi-level governance arrangement, involving hierarchical regulation by both national governments and UNFCCC administrative bodies (Bumpus and Liverman 2008; Stripple 2010). Overarching authority rests with the CDM Executive Board, which is responsible for developing methodologies, regulating the project accreditation process and issuing Certified Emissions Reductions (CERs). The CDM also involves bottom up project design and innovation by non-state actors, often in partnership with private industry. This suggests both a convergence of CDM governance processes around international rules and market operation, and a divergence of governance modes at the national and project level based on domestic institutions and local political factors (Newell 2009). The proliferation of different initiatives, regimes and processes has led some authors (Biermann et al. 2009; Zelli and Van Assel 2013) to argue that climate change governance is fragmented and diversified across an array of institutions that shape policy responses at different scales. Fragmentation and diversification recognise that climate change is marked by a collection of international institutions that are different in organisation and norms, have multiple public and private constituencies, and a spatial scope that extends across national, bilateral and international domains (Biermann et al. 2009; Zelli and Van Assel 2013). Diversification also occurs in actor relationships and discourses that dissect political decision making and lead to novel and new institutional arrangements. Whist there is clearly a level of institutional fragmentation and diversification, the approach identifies different applications of ostensibly similar governance processes and policy instruments. Since the introduction of the Kyoto Protocol, the global and national approaches to climate mitigation have been dominated by market instruments, and a series of voluntary measures that share a level of commonality across jurisdictions. Whilst the application of carbon markets may vary and adapt to the national or local context, the broader governance rationale that underpins these instruments bears many similarities. For example, approaches to measure, report and account for carbon share similar theoretical, technical and institutional context, driven by common rules and standards, and by corresponding knowledge systems and administrative structures. 46

67 2.8.2 REDD+ Architecture Research on REDD+ is subject to various governance approaches or theoretical orientations. A significant proportion of the literature adopts the concept that REDD+ governance is shaped by the relatively fixed and territorially bounded spheres of authority, and notably the international framework under the UNFCCC, state level processes and sub-state activities. The literature seeks at address governance arrangements or interplay across these political and institutional levels, often to evaluate how international initiatives and policies diffuse into government programs or local initiatives. As Corbera and Schroeder (2011, pp.91-92) argue: interplay may occur vertically between the top-down international REDD+ norms and procedures and existing national-level policies and local-level institutions (e.g. property rights) and pilot activities, or horizontally between national PAMs [policies and measures] aimed at reducing deforestation and degradation and other national land-use related policies and processes. Institutional structure is therefore a common theme to emerge from these REDD+ governance approaches, in terms of the design and operation of various rule systems and policy measures, and the administrative systems charged with implementing and operating them. As part of this approach, research attempts to evaluate different components or elements that shape REDD+, in order to identify options to improve forest and carbon governance at different scales. There are several approaches to visualising and describing REDD+ governance. One approach focuses on architecture and relates to the institutions, organisations, principles, norms, mechanisms and decision-making procedures (Vatn and Angelsen 2009; Biermann et al. 2011; Corbera and Schroeder 2011). Architecture includes those institutional features that shape activities around forest and land management, such as the international and national funding and payment distribution mechanisms (Vatn and Vedeld 2013). In comparison, carbon markets engage innovative actor networks who establish new institutional nodes along the carbon commodity value chain; with each node servicing a specific function to produce, certify, accredit and trade in forest credits (Hajek et al. 2011). Analyses of governance architectures often point to their multi-level nature, focussing on the actor relationships and institutional linkages between local scale projects, national programs and international rules and standards (Skutsch and Van Laake 2009; Mwangi and Wardell 2012). The multi-level 47

68 approach is applied to the analysis of REDD+ program areas such as finance and benefit distribution mechanisms, MRV and accounting systems (Korhonen-Kurki et al. 2013), rules and processes of forest tenure (Doherty and Schroeder 2011), and multiactor relationships across networks or organisational levels (Forsyth 2009). A variation is the nested governance approach, focussed on embedding localised incentives, practices and actors within higher scale processes of national and international REDD+ (Pedroni et al. 2009; Agrawal et al. 2011; Kashwan and Holahan 2014). Another approach to REDD+ governance focuses on accountability and legitimacy, applying a range of evaluative criteria to determine the quality and effectiveness of institutions and political processes. As discussed above, questions on governance legitimacy relate to whether the various public policies, and private run programs actually deliver on the broader claims about climate change mitigation, but also on issues of stakeholder engagement, equity, and benefits. Legitimacy is therefore a performance based approach to REDD+ governance. Legitimacy can relate to the justification and acceptance of political authority, where a legitimate institution has a right to govern on the basis of expertise or public accountability (Bodansky 1999). Legitimacy theories have several interrelated elements. Input legitimacy is based on the source of political authority and procedural qualities, such as democratic accountability and transparency in decision making; and output legitimacy, or the effectiveness of measures to solve the problems as prescribed (Lövbrand et al. 2009a; Biermann and Gupta 2011; Karlsson-Vinkhuyzen and McGee 2013). In the case of REDD+, participatory institutions are considered an important element of input legitimacy and the quality of governance processes, ensuring that multiple stakeholders are part of national and international activities (Cadman and Maraseni 2012, p.632). Several studies evaluate the input/output legitimacy of different REDD+ governance institutions. Lederer (2011) applies an input/output legitimacy framework to evaluate the institutions and processes of the CDM and to compare this with REDD+. Both the CDM and REDD+ do provide a level of input legitimacy in terms of formal engagement processes, but the CDM is designed for greater output legitimacy in terms of emissions reductions. Rosendal and Andresen (2011) look at options for an international REDD+ financial mechanism based on the Global Environment Facility (GEF). Criteria for input legitimacy include appropriateness and fiscal accountability of various mechanisms, whilst outputs are cost efficiency, operational effectiveness and capacity to deliver. The authors (Rosendal and Andresen 2011, p.1914) argue that the GEF provides 48

69 accountability for payments and also allows integration of social and biodiversity priorities into payment options. Vatn and Vedeld (2013) use a similar framework to evaluate different funding instruments in shaping REDD+ governance at the national level. The evaluation covers four institutional designs, including market and project based approaches, and compares this with state operated mechanisms, such as a stand-alone fund, or budget support. They (Vatn and Vedeld 2013, p.430) reach similar conclusions that public funds provide greater legitimacy, accountability and transparency, whereas the market option is the lowest in terms of output legitimacy, with limited potential to produce tangible outcomes in developing countries Agency and Discourse Agency is a further theme in REDD+ governance, and particularly the capacity and authority of state and non-state actors to design and operate institutions and processes. Agency in climate and environmental governance has been described as the purposeful steering by multiple state and non-state interests, either directly by making decisions or indirectly by shaping the preferences and actions of others (Schroeder 2010; Biermann et al. 2011). For environmental and climate change governance, the expanding agency of non-state actors is a defining feature of international and domestic programs (Bulkeley and Newell 2010). Multiple actors from NGOs, businesses, communities, as well as influential individuals provide input on the international and national policy processes, shape the operating rules and standards, and conduct local scale project activities. However, critical questions remain about the relationship between structure and agency, and who bears authority in the shifting arrangements of REDD+ governance. Schroeder (2010), for example, takes the case of indigenous groups in the international REDD+ process to argue that although they remain engaged in the process through formal and informal means, their agency remains relatively weak and dependent on the support offered by participating states. Discourse is an important analytical tool to understand how REDD+ governance processes unfold. There is a trend towards discursive-institutional approaches in forest governance aimed at understanding relationships between policy design, discourse and power (Arts and Buizer 2009; den Besten et al. 2014). These approaches investigate the nature of structure and agency and the capacity of different actor networks or coalitions to utilise discourse in order to influence the design and also implementation of REDD+ policy frameworks (Di Gregorio et al. 2012; Brockhaus et al. 2013). Discursive strategies in this context are a type of mechanism that enables 49

70 ideas and knowledge to change, and to either limit or expand the institutional structures of forest governance (den Besten et al. 2014, p.46). Empirical research by Brockhaus et al. (2013) on REDD+ programs in six countries investigated the relationship between policy actors, the institutions within which they operate, and the discursive and organisational mechanisms they employ. They argue that transformational change to forestry practices in most countries will be constrained by economic path dependences, often supported by influential resource networks. Change requires coalitions of NGO interests to align with state agencies in order to advance discourses around sustainable forest management, indigenous rights and carbon mitigation (Brockhaus et al. 2013, p.7) Good Governance REDD+ institutional reforms at the national and sub-national level often approach governance as a normative goal. Rather than a focus on governance though agency or steering, these approaches seek to evaluate and identify reforms and practices that can improve the public administration and accountability around forest resources. Since the 1990s, international forestry policy has aimed to improve the way developing countries manage forests through a number of programs such as sustainable management, good governance and decentralisation (Agrawal et al. 2008; Gupta 2012; McDermott 2014). Many developing countries have initiated decentralised governance in order to transfer authority and capabilities from central administrations to the local or district level: primarily to improve efficiency by redistributing resources and benefits, but also to increase accountability to local stakeholders (Tacconi 2007b; Larson and Soto 2008; Petkova et al. 2010). Recent reform approaches under the banner of good governance focus on measures and indicators such as accountability and transparency, democratic processes, the rule of law, political participation and stability of state institutions (Kaufman et al. 2011). Research on governance quality in REDD+ advocates the importance of national and sub-national indictors such as fair decision making, transparent and accountable forest administration and financing, equitable tenure and land reform processes, and policy coordination across administrative levels (Larson and Petkova 2011; Pettenella and Broto 2012; Fosci 2012). This research also identifies the significant challenges to implementing REDD+ in countries with weak public sector institutions, and where political instability and corrupt or ineffective judicial systems affect all levels of government (Larson and Petkova 2011; Karsenty and Ongolo 2012; McDermott 2013). 50

71 Recent studies also assess forest governance to determine what is required for successful implementation of REDD+ (for example Kanowski et al. 2011; Cronkelton et al. 2011). These studies essentially attempt to quantify and measure forest governance qualities or features to inform policy processes. Cross-country research has focused on the institutional dimensions that influence the direction and potential success of national REDD+ processes (Di Gregorio et al. 2012; Murdiyarso et al. 2012; Korhonen Kurki et al. 2013b). Countries which have made progress towards a national REDD+ framework, such as Indonesia, Brazil and Vietnam often have an institutional setting in place, in terms of existing policies and adequate administrative capacity to conduct reforms, in addition to supportive constituencies (Korhonen Kurki et al. 2013b, p.182). Murdiyarso et al. (2012, pp ) argue, however, that apart from Brazil, most developing countries lack the institutional capacity or transparency to implement effective and efficient policies that address the drivers of deforestation, or indeed implement REDD+ programs. Studies also evaluate governance conditions that will enable project implementation. For example, Petenella and Brotto (2012) evaluate the governance conditions for successful carbon offsets in Peru pointing to sound land-use planning and allocation frameworks, combined with public institutions that support local communities to manage forests. 2.9 Limitations with Governance Approaches Studies utilising various governance approaches offer important insights on how REDD+ is designed across scales, which agents participate and exercise authority in different contexts, and what attributes make some programs and projects successful. Governance theories also provide a valuable framework to analyse the relationships between institutional change, multi-actor agency, and discourse. Critically, however, there are some important limitations with governance theories and their current application to climate change and REDD+. To start, governance research often adopts fixed notions of the nation state as a relatively stable entity operating within an interstate regime, but with a redistribution of power to some non-state actors (Sending and Neuman 2006, p.652). As Okereke et al. (2009, pp.65-68) argue, the state is still considered to have ultimate sovereign authority in climate policy, producing an often fixed and rigid conceptualisation of state functions. Authority to implement REDD+ activities is therefore largely dependent on public institutions. Whilst governance approaches recognise the increasing agency of non-state interests they often relegate 51

72 their role as secondary to the state, rather than paying closer attention to the myriad of ways that various civil society and private interests actively create and govern the climate change domain (Bulkeley et al. 2012). Non-state interests, including organisations, businesses, households and individuals play a critical role in conducting the myriad of REDD+ activities, and in governing forest economies and environments more generally, and this requires further detailed evaluation. Governance theories are often attendant to the rise of markets and privatisation or contracting-out of government functions. As such there is research on the establishment and operation of environmental markets and related private institutions, and the effects on different local stakeholders and forest resources. Research on REDD+ has also analysed MRV, tenure, finance and benefit distribution, suggesting that governance is not necessarily conditional on redistributing authority to privatised market spaces, but encompasses a whole range of policy mechanisms and processes. However, governance approaches provide few analytic insights on the various regulatory, planning, legal and technical approaches that remain fundamental to administering environmental resources. Indeed, governance theories largely ignore how public agencies continue to employ multiple non-market policy instruments to REDD+ and their administration through large centralised and multilayered bureaucracies. There is also lack of attention to the function of knowledge and expertise in governance approaches, and importantly the relationship between knowledge, particular policy instruments and the methods employed in governing. Governance research applied to climate change and REDD+ often treat power as a commodity used to influence people or entities, either through institutional arrangements and decision-making, or by indirectly shaping the preferences of others. In the analysis of REDD+, power is viewed as something held by certain interests, to be harnessed or exercised to achieve certain ends in terms of delivering initiatives, or in obstructing and limiting reforms. In the local setting power is conceptualised as the ability of one actor to influence others through coercion, incentives and information, and thereby shape governance structures (Krott et al. 2013). Power is a zero sum exchange where one actor s gain out of a process is another s loss (Okereke et al. 2009, p.62). Power is therefore relative to the type of actor, their relationship to particular institutions of government, the market or civil society, and their capacity to shape institutional design through discourse, networking and other strategies (Brockhaus et al. 2013). Critically, governance accounts that ascribe power as something to be exercised over others fail to address the constitutive and empowering 52

73 effects of actor relationships, or how power is produced in various contexts and settings. The concept of governance as steering, suggests that power is produced through the interactions and discursive exchanges between actors that occur in the development and operation of multiple activities. The governance literature on REDD+ also pays significant attention to the structural design and mapping of governance architecture. The literature seeks to analyse institutional quality and legitimacy in terms of delivering outcomes, and meeting the perceived needs of various stakeholders. In developing countries this often focuses on creating and assessing the conditions for effective government, in terms of building public sector capacity and transplanting (essentially western) democratic governance arrangements to manage forests and land. Whilst this research is important, it provides limited insights on the implementation of REDD+ components in developing countries with vastly different political systems, and complex social and cultural practices around forests. Further, governance approaches often identify the policy instruments, norms, processes and organisational arrangements, but provide little analysis into how these elements individually and collectively operate to govern the environment. Indeed, current analyses pay limited attention to how the different elements function, in terms of what particular policy mechanisms and processes seek to achieve, which actors utilise them and why, who engages in REDD+ and for what ends, and how the multiple elements relate to one another in a larger governing framework Further Research The flurry of REDD+ activities across numerous organisations and countries has resulted in an enormous level of complexity, fragmented decision-making and ambiguity (McDermott et al. 2014). Moreover it is unclear how the suite of national policy reforms, demonstration activities, safeguards and technical upgrades will transform and realign the governance of tropical forests to focus on carbon and GHG mitigation. One of the more significant gaps in the research is detailed empirical analysis of the major national level processes and their relationship to sub-national or pilot program activities. To understand how REDD+ is governed requires in-depth theoretical analysis of the multiple institutions and processes, and the interaction across jurisdictions and between agencies and organisations. In addition, there is a need to examine how governments and other organisations apply these different policy mechanisms to REDD+ in specific countries, and what factors inform their design and 53

74 implementation. Expanding the analysis to examine how the combination of policy instruments and processes operate as a system or regime of governing is important for understanding how REDD+ could potentially transform forest governance in developing countries. Detailed empirical research is required into the design of REDD+ national strategies and sub-national programs, and how they integrate or conflict with existing economic development, forestry and land-sue planning frameworks in particular countries. There is the need to understand existing synergies and contradictions across policy areas at national and sub-national levels, including conservation strategies, and agricultural and urban development programs (Corbera and Schroeder 2011, p.96). This research should explore the problems of deforestation in particular countries, and how this shapes existing and future policies. Case based research also needs to account for the challenges of policy implementation and enforcement within countries with limited resources and capacity, and high levels of illegality and political corruption. Understanding the interplay between existing forestry polices, political institutions of the state, and the existing resource regimes will illuminate the barriers to effective implementation. Situating REDD+ research within broader political and socioeconomic contexts, including democratic and non-democratic regime issues, will also enhance an evaluation of why governments implemented certain policies, and why others face substantial challenges. As outlined above, there are limitations with the governance approaches. Understanding how REDD+ operates and is governed requires suitable theoretical frameworks that attend to the complexity and variance in the activities and processes underway. Research is required to develop a theoretical framework that can be systemically applied to the combination of REDD+ initiatives and activities in specific countries. This theoretical framework needs to account for how power functions in different contexts, and the ways in which different actors engage in the operation of projects, and in broader policy reforms and climate programs. This research needs to expand from a focus on governance modes or institutional architecture and take account of the various forms of knowledge and technical systems that inform and shape REDD+. A theoretical framework also needs to address broader systems or regimes of government within which REDD+ is embedded, particularly in developing countries that struggle with democracy and government accountability. In this regard, Chapter 4 introduces the theoretical framework used for this thesis. The next Chapter (3) outlines the methodology and the selection of Indonesia as the case study. 54

75 Chapter 3: Research Methodology 3.0 Introduction Chapter 3 presents the methodology applied to this thesis. The review in Chapter 2 explored the multiple and overlapping policy approaches, technical initiatives and activities that comprise the program Reducing Emissions from Deforestation and Forest Degradation (REDD+), and various governance approaches used to analyse these reforms. This Chapter provides an overview of the research orientation, in addition to the methods for data collection and analysis, and the adoption of a single embedded case study design. Importantly, the Chapter justifies the selection of Indonesia as a case study to analyse the combination of REDD+ policy initiatives and activities in one country. Indonesia was selected because the Government of Indonesia (GoI) and partners are conducting one of largest REDD+ programs, with initiatives spanning the national and sub-national scales. Indonesia therefore suits the research objectives and offers a critical case to apply the theoretical framework. To conduct the case study, the research applies methods of qualitative data collection and analysis, including semi-structured interviews combined with document analysis. Qualitative methods enable an in-depth analysis of the multiple REDD+ components and support an interpretive approach to policy analysis, where the focus is on the views, discourses and interpretations of the actors involved. This Chapter is structured in three sections. The first section focuses on the research orientation and the application of qualitative methods. The second section covers the research design, whilst the last section outlines the data collection and analysis methods, and discusses sampling and the ethical conduct of the research. 3.1 Ontology and Epistemology Ontology and epistemology are important for shaping the methods and theory used by social science researchers. Ontology relates to the claims or assumptions that a particular approach to social enquiry makes about the nature or form of reality to be investigated (Hay 2002; Furlong and Marsh 2010; Lincoln et al. 2011). Epistemology relates to what can be known about the world and how knowledge is produced in the 55

76 research process. Epistemology and ontology are therefore invariably linked in social enquiry. Furlong and Marsh (2010, pp ) identify two broad ontological positions in political science - foundationalism and anti-foundationalism. Foundationalism comprises positivism and realism and contends that a real world exists separate from the knowledge of the researcher. For positivists (and related post-positivists) research is an objective exercise that employs empirical quantitative analysis and the testing of causal hypotheses and replicable causal explanations (Moses and Knutzen 2007, pp.42-43). The second foundational ontology of realism accepts that structural relationships exist between phenomena. However, these real world structures often remain obscured and are difficult to observe, and rather than determining outcomes they mediate social behaviour through ideas (Hay 2002). Realism is also willing to accept quantitative and qualitative analysis in social science, and that the interpretation by different actors can affect outcomes and change structures (Furlong and Marsh 2010, p.205). Some critics argue (Fischer 2003, pp ) that positivist and realist approaches have come to dominate public policy analysis, yet they fail to explain fully the values, intentions and competing claims behind issue formation and social action. The ontology of anti-foundationalism includes various orientations, most notably constructivism, and related approaches such as interpretivism and postmodernism. Anti-foundationalism broadly argues that reality is socially constructed, and that a real world does not exits beyond a person or group s knowledge of it (Furlong and Marsh 2010, p.190). Constructivism rests on the philosophical assumption that social agents construct understandings about the nature of the social and physical world, such as through ideas, beliefs, norms and identities (Parson 2010; Lincoln et al. 2011). People can construct different perceptions of reality, but social, political and cultural processes inform these perceptions. The constructivist approach also contends that a researcher cannot study social phenomena separately from any interpretation of them, and that these interpretations are not objective but shaped by social and discursive processes. The focus of social science research is thus on the interpretation of meaning rather than a search for causal mechanisms. Further, knowledge is not used explicitly to verify or falsify theoretical propositions but supports arguments in the ongoing debate about the most relevant theoretical framework for understanding and constructing social reality (Hay 2002). 56

77 The modernist stream of anti-foundationalism would accept that components of the world could be determined through careful and subjective analysis of the knowledge obtained from different social agents (Parson 2010, p.91). Knowledge is, in part, the product of social negotiation or collective reconstructions, often leading to some level of consensus amongst people or groups. Post-modernists on the other hand, argue that it is not possible to produce commonly shared understandings or accepted truths about the social world (Moses and Knutzen 2007; Furlong and Marsh 2010). Instead, divergent theoretical frameworks lead to contending interpretations or discourses of social reality. For post-modernists, notions of reality and knowledge are inseparable and reproduced through relations of power. The key aspect to the interpretive approach rests on a philosophical analysis of meaning in action, and this requires understanding the beliefs and ideas embodied in certain practices (Bevir and Rhodes 2005, p.174). In a public policy context, interpretive approaches attend to the views and interpretation that different actors bring to certain issues, and how the policy process can generate new meaning and understanding (Yanow 2007, pp ). The interpretive approach thus focuses on how different parties frame policy options and any potential effects, and how these collective beliefs inform policy acts such as legislation, but also broader political processes and social reforms (Yanow 2000, p.12). The aim of interpretation is to identify consensus or agreement amongst a group or community engaged in a particular policy program. Discourse is also important for understanding how policy is shaped and influenced through meaning and ideas (Bevir and Rhodes 2005). Different groups may develop a collective understanding of a particular issue or policy, and these are often expressed through common language, discourses and debates between rival perspectives. Interpretive analysis also focuses on the intention of particular policy undertakings or actions, implying that a researcher can identify the future state of affairs that required the certain action, rather than identify or test a specific causal relationship. This research adopts the ontology of anti-foundationalism and associated constructivist or interpretive approach to policy analysis. The researcher does accept a realist component to the topic, in that deforestation and climate change are physical phenomena that have tangible social and environmental impacts. However, this research project is concerned with understanding the different political and social responses to deforestation and how various public and private agents construct, conduct and implement the multiple REDD+ activities. This research also aims to 57

78 uncover the intention or objective of a particular policy reform and program activity, and how these policy instruments function to change forestry and land-use practices. An interpretive approach therefore suited the research, as the aim is to understand how different actors involved in REDD+ interpret and collectively view the different policy and program activities. An interpretive epistemology also facilitates an analysis of the different perspectives that social agents adopt in relation to forests and climate change, and importantly how they construct and shape policy responses (Boyd 2010). The different policies and processes are shaped by contending theoretical assumptions, and utilise various forms of knowledge. The design and implementation of REDD+ is also framed within ongoing discourses around the cultural and economic value of forestry and agriculture to society. Importantly, a constructivist approach enables an analysis of these different ideas, theories and discourses by incorporating the subjective interpretations of multiple stakeholders. These multiple ideas on REDD+ can be found in government departments, rural communities, civil society, external organisations and business interests who express both complementary and competing arguments about deforestation and how governments should respond Qualitative Research The research project adopts a qualitative approach to research, which is consistent with a constructivist ontology and epistemology. Qualitative research can be defined as an interpretive and naturalistic method that aims to understand social phenomena through the different meanings that people ascribe to these phenomena (Patton 2002; Stake 2010; Denzin and Lincoln 2011). The research is often conducted in the setting or location where participants are regularly engaged in the relevant activities. Qualitative research is therefore conducted through interaction with different actors to determine how they understand and shape the world around them (Tracy 2012, pp ). Quantitative research, in comparison, aims to measure and calculate certain attributes or variables across a population or social group (Seale 2012). The qualitative approach suited the research project as it aims to understand how REDD+ policies and programs are conducted and governed in a specific country, with an emphasis on the domestic conditions that shape the processes of change. This enables the research to investigate how different stakeholders and practitioners understand and engage in the different organisational reforms and political processes underway. 58

79 A qualitative design and methods also facilitated in-depth research on the topic. Qualitative research often focuses on the development of rich and detailed descriptions of the subject matter (Denzin and Lincoln 2011, p.5), and aims to generate a holistic understanding of social phenomena and the complex interdependence of issues (Vromen 2010, p.257). Context, such as the particular social and historical period or setting is important in data collection and analysis (Stake 2010; Tracy 2013). A qualitative design suited the research as it aims to provide a detailed and rich description of the overlapping series of policy and technical reforms that span the international, national and sub-national scales. The research will also situate the study on REDD+ within the political, economic and cultural traditions that characterise forest and land management in Indonesia. Further, the study will assess the different ways that deforestation is discursively framed and how this relates to the types of programs and projects deployed. As this is a contemporary topic, the majority of the qualitative field research will be conducted in the country where the programs are underway. 3.2 Case Study Design The research employs a case study method. Case studies are widely used in policy and governance studies but there are contrasting definitions and applications of this methodology. Gerring (2007, p.19) defines a case study as the investigation of a spatially contained or bounded phenomena, or unit of analysis, that covers a defined temporal period. According to Yin (2009, p.18) case studies can also be characterised as studies that cover a contemporary phenomenon involving a number of variables. In policy analysis, case studies can cover the nation state, a government agency or nonstate organisation, or alternatively the unit of analysis may be a specific policy program. Case study design varies and includes in-depth single or small N case studies that aim to understand complex phenomena through detailed empirical analysis (Gerring 2007; Yin 2009). In contrast, multiple or larger N designs aim to replicate or compare several variables across a number of cases by using a narrower or thinner set of empirical data. Alternative methodological approaches, such as surveys, would use statistical analysis to investigate a few variables across a large population or representative sample (Seale 2012). Theory is important in the design of case study research and informs data collection and analysis (Yin 2003, pp.38-41). Case studies can be used to test existing theories or to generate new hypotheses and theoretical principles (George and Bennet 2005; 59

80 Moses and Knutsen 2007). The comparative method in political analysis involves the development and testing of theories and causal relationships across cases, often as a large N or small N comparison between countries (Hopkin 2010, p.286). In contrast, a single or small number of in-depth case studies can apply abstract theoretical principles to interpret and explore complex social problems (Blatter and Haveland 2008, pp.21-22). Single case designs can also cover variables that best represent the theoretical concepts under analysis, achieving high levels of conceptual validity (Yin 2003). However, there are drawbacks with single and small N case studies. They are often criticised for their lack of generalisability or external validity, which can limit the application of findings to other locations or broader social processes (Flyvbjerg 2011, pp ). Single cases are challenged on the grounds of bias and reliability, with case selection designed to confirm or disprove particular theoretical orientations, or the opinions of the researcher (Yin 2009; Flyvbjerg 2011). These limitations mean that case study design and selection is critical. The research has adopted a single embedded case study design that was informed by the research questions and aims. There are approximately 60 countries engaged at some level in REDD+, providing significant opportunity for either a large or small N comparative analysis. A comparative analysis across several countries would permit the study of several key variables, providing a less in-depth analysis but a focussed testing of theoretical concepts (George and Bennet 2005, p.82). Although a comparative analysis has many benefits, it was not selected because the research aimed to analyse the multiple components or variables to REDD+ in a specific country. A single case study enables a comprehensive study of the suite of programs, policies and projects in order to unravel the complex thinking and organisation behind these initiatives. A single in depth study therefore suits the research questions, which aim to understand how this complex governing regime is designed and how it operates in Indonesia, including the barriers and limitations. The research used an embedded case study model because it facilitates analysis of the complement of REDD+ programs and initiatives across different scales. An embedded case study is structured as a larger study with a number of sub-units of analysis (Yin 2009, pp.42-43). For this thesis the case will incorporate the reform process across national, provincial, and district jurisdictions (see figure 3.1). At the national level the sub-units include; the major policy and administrative changes; the establishment of financing and benefit distribution mechanisms; capacity building programs and technology development, such as measurement, monitoring, reporting, 60

81 and verification (MRV) systems; and participation and consultation strategies. At the sub-national level, the units of study will cover provincial pilot programs and local demonstration projects, with a focus on benefit distribution, safeguards, tenure issues and stakeholder engagement. Further, at the local scale both market and non-market based demonstration projects will be selected to represent the theoretically different approaches to mitigating forest emissions. REDD+ National Policies, measures and activities Sub-national Provincial pilot activities Projects Market/incentive Non-market Figure 3.1 Embedded Case Study Design Selection of Indonesia There are a number of potential countries that could be selected for this research. Countries including Brazil, Peru, Vietnam, Indonesia, Central African Republic and Tanzania have all made progress on REDD+ at the national and project scale. A comparison of possible case studies is made in Table 3.1. Indonesia is one of the early global leaders on REDD+ and has made progress on key components, such as a national strategy and MRV system, and is host to approximately 44 demonstration projects. The sheer scope of Indonesia s reform agenda provides a significant opportunity to undertake an in-depth analysis of the design and implementation of one of world s largest REDD+ initiatives. Indonesia offers a comprehensive range of variables or components across political sub-units or scales. A case study of Indonesia therefore suits the research problem and objectives in that the government 61

82 and partners have initiated an extensive and evolving multi-level REDD+ program. Indonesia also ranks as one of the top five countries in terms of greenhouse gas (GHG) emissions and has one of the highest domestic rates of tropical deforestation, and is therefore critical to the global response to REDD+ (Irawan et al. 2013). Data availability and access was also a major consideration in the selection of the case study (Blatter and Haveland 2012, pp.30-32). Indonesia offered access to a large volume of data and information on REDD+, generated by the numerous government agencies, non-government organisations (NGOs) and other interest groups. Table 3.1 Progress in REDD+ Countries Country Deforestation National Dem Sub- Government rates strategy Project national type Indonesia Very high Complete 44 Progress Democracy (transition) Brazil Very High Incomplete 53 Progress Democracy Democratic High Incomplete 11 Some Authoritarian Republic of Congo Progress Tanzania High Incomplete 10 Some Democracy Progress Vietnam Low Complete 7 Limited Authoritarian PNG Low Not started 4 Limited Democracy (CIFOR 2014) In addition, Indonesia affords the opportunity to investigate the significant challenges to implementing REDD+ in a rapidly developing country. This addresses a current gap in the literature, the need to evaluate domestic issues that impact on national and subnational programs. A critical issue for REDD+ is the extent to which governments are capable and willing to overhaul forest management, particularly in countries geared to resource development and with influential industry sectors. In Indonesia the forestry, mining and agricultural industries are integrated with political authority, and dominate politics at all levels of government (Gellert 2010). Further, Indonesia is undergoing a difficult transition from an authoritarian regime to a democratic polity, and whilst the country has improved its governance indicators there remain systemic problems in the resource sectors (McCarthy 2011; Hamid 2012; Fukuoka 2013). REDD+ reforms face 62

83 challenges in the context of a rapidly developing economy, entrenched corruption within state institutions, coupled with limited capacity to implement an international climate and forests program. To provide a working example of how an incentives scheme and carbon market will operate in Indonesia the thesis will provide a sub-unit analysis of the Kalimantan Forest and Climate Partnership (KFCP), a REDD+ demonstration project located in Kapaus District in Central Kalimantan Province. The KFCP was selected because at the time of research it provided one of the more advanced demonstration projects focussed on developing and testing different components for a carbon market. 9 The project displays a number of factors that characterise the major challenges with implementing REDD+ markets and incentives including; competing land-use and tenure claims; relatively limited infrastructure and development opportunities for communities; and issues over corruption, accountability and enforcement. The KFCP thus provides a representative unit of study at the local project scale. Central Sulawesi was selected as a sub-unit study to analyse the introduction of participatory and engagement strategies. Central Sulawesi is a provincial scale REDD+ demonstration program setup to develop the policy, institutional and technical framework. Although there are a number of pilot provinces, Central Sulawesi was selected because different agencies had trialled a range of participatory processes including consultation forums and free prior and informed consent (FPIC). This provided the opportunity to evaluate how deliberative strategies were conducted, their effect on the REDD+ policy process, the stakeholders involved, what the processes aimed to achieve, and importantly any limitations. Central Sulawesi offers a representative unit of analysis in terms of provincial scale activities. The selection of Indonesia also affords the opportunity to apply and test theories of governmentality to climate policy in a developing country. Single cases are often used when applying an existing and well developed theory to a new context and set of phenomena (George and Bennet 2005; Gerring 2007; Yin 2009). The Indonesia case study is significant because it provides one of the first applications of governmentality theory to REDD+ (see Gupta et al. 2012; Boer 2013; Stephen 2013; Stephen et al. 2013), and importantly the combined theoretical framework developed in this thesis 9 At the time of research the KFCP had progressed through the design phase towards implementation of key project activities, whereas a number of other local demonstration projects remained at the conception or early design phase. 63

84 (and elaborated further in Chapter 4). The combined theoretical framework will be used to analyse the multiple policies and initiatives that function to change forest and land-use management. However, many of the programs and initiatives underway in Indonesia are similar to those in the majority of countries undertaking REDD+. Indonesia therefore, provides a partial and yet critical test of the theoretical framework. The thesis will therefore evaluate the utility of the combined governmentality framework for interpreting REDD+ in Indonesia. The Chapter now turns to data collection methods. 3.3 Data Collection and Analysis Data collection involved qualitative interviews combined with extensive document analysis. Qualitative interviews offered a number of benefits for the research. In particular, they offered a degree of depth and complexity in the data that is often difficult to obtain from other methods such as questionnaire surveys (Rubin and Rubin 2005; Yin 2009; Byrne 2012). Qualitative interviews involve an informed dialogue between the researcher and the participants and thus generate knowledge through social interaction (King and Horrocks 2010; Roulston 2010). A further strength of qualitative interviews is the ability to focus on the attitudes and interpretations of different respondents, and to probe complex relationships. 10 In policy studies, qualitative interviews can provide different assessments of the political process, or the ideational factors behind agency and decision making, such as norms, values, learning and cognition (Rathbun 2009). A rationale for the selection of qualitative interviews was the need for rich and detailed information on program and project activities, and to investigate the various insights, interpretations and reasoning of key respondents. This is important, because REDD+ is a highly contested political domain subject to competing ideas and values regarding suitable approaches to manage forests and carbon, and the impacts and benefits of different initiatives Semi-structured Interviews The researcher used a semi-structured interview format. Semi-structured interviews involve a research guide, with a set of broader questions followed by secondary 10 Constructivist approaches suggest that the analysis of interviews can also take into account the specific situations or personal accounts of each interviewee, in addition to the interviewer. This aspect was not explicitly explored by the research, however in the following Chapters there is regular discussion of particular interviewee responses, including the use of in-text quotations. 64

85 questions and probes (Patton 2002, p.349; Roulston 2010, p.15). This format offers an overall structure to cover key areas of the research topic, combined with flexibility to explore new issues that arise during the interviews (Rathbun 2009). This was more applicable than structured interviews that may have proven inflexible to new topics, or open ended interviews that may have missed important areas of questioning. Interview questions were informed by the research questions and the theoretical framework (Kvale 2007), and developed from key literature in the field, including primary documents from Indonesia and other countries. The questions aimed to understand relationships between different programs, the interplay between local, national and international rules and policies, and the role of different state and non-state actors. In this regard, the same questions were asked of each interviewee but with a greater focus on their areas of expertise and knowledge. A copy of the interview question sheet is located in Appendix A. The researcher used a combination of purposive and snowball sampling to recruit suitable interview respondents (Patton 2002). The overall aim was to recruit a crosssection of professional practitioners engaged directly in REDD+. In the first stage the researcher used purposive sampling, and sent recruitment letters to approximately 50 agencies, organisations and companies working on REDD+ in Indonesia. When fieldwork commenced the researcher used snowball sampling and asked interviewees to recommend other people. The researcher conducted 35 interviews with respondents from 22 agencies and organisations in Indonesia and several in the Southeast Asia region. A list of interviewees and organisations is provided in Table 3.2 below. The majority (30) of the interviews were in-person, with the remaining conducted by telephone or Skype - which tended to produce shorter interviews and less dialogue, a commonly cited drawback (Rubin and Rubin 2005). All interviews were conducted in English, and recorded with an audio device and written notes. A data company transcribed the interviews. 65

86 Table 3.2 Interviewees and Organisations Agency/organisation Position Government of Indonesia Multilateral agencies Bilateral agencies International environment NGOs Indonesian NGOs Researchers and private consultants National REDD+ Taskforce (x1) National Climate Change Council (x3) President s Office (x1) Ministry of Foreign Affairs (x1) Ministry of Forestry (MoF) (x2) Ministry of Planning (BAPPENAS) (x2) Ministry of Environment (MoE) (x1) UN-REDD+ Programme Indonesia (x3) UN-REDD Programme Southeast Asia (x1) United Nations Environment Program Indonesia (UNEP) (x1) Indonesia/Australia Forest Carbon Partnership: AusAid (IAFCP) (x3) FORCLIME and GIZ German Forests and Climate Change Programme Indonesia (x2) The Nature Conservancy (x3) World Wide Fund for Nature (WWF) (x2) Fauna and Flora International (FFI) (x1) Pact International (x1) WAHLI/Friends of Earth Indonesia (x1) HuMa (Association for Community Based and Ecological Law Reform) (x1) Carbon forestry consultants (x2) Academic institutions (x3) National director Program director Program managers Project manager Senior director Senior directors Program directors Policy officer National director Pilot province director Technical manager Regional director Project manager Program specialists Project officer National director Project manager Regional director (SE Asia) Program manger Project officer Program director Policy advisor Technical officer Project manager Campaign manager National director Project consultant Senior researcher 66

87 There were several challenges with the sampling and interview process. The researcher sent recruitment letters to industry groups, and to private companies with interests in REDD+ projects. Unfortunately there were few responses from industry groups. To address this gap, data was sourced from on-line materials such as websites and public submissions. Fieldwork was originally planned for Central Kalimantan to interview representatives from local communities and provincial or district governments. However, Indonesian research visa conditions restricted research to West Java Province only and prevented the researcher from travelling to Central Kalimantan and conducting interviews (see explanation below). Further, the primary researcher became seriously ill and had to return to Australia early. Although the researcher gathered a large and high quality body of data, the limited number of interviews with representatives from industry and communities was a drawback and reduced the depth of information on local carbon projects. The analysis therefore relied on the 10 interviews with project managers and NGOs, and on documents and secondary reports Documents Documents offered the other major source of data. Documents are particularly useful for in-depth case study observations in qualitative research (Yin 2009). Documents provide a background to an issue that enables the researcher to interpret and generate meaning from social processes (McCulloch 2004; Bowen 2009). Through different forms of language, documents provide a discursive representation of the ideas, concepts and ideologies that shape social action, and are therefore intimately linked to the way knowledge is generated and to the exercise of power. For this reason, documentary evidence requires critical evaluation or reading to determine the context in which it is created, the intended audience, and the orientation of the authors (Bowen 2009). As McCulloch (2004, pp.44-46) argues, utilising multiple documentary sources is an important method for generating a critical interpretation of an issue area. Adopting this critical approach the researcher collected a large volume of primary and secondary documents from a cross section of state agencies, civil society organisations, community groups and private industry. The types of documents collected can be found in Table 3.3 below. A critical issue is the authenticity or reliability of the document, which is often context dependent (McCulloch 2004; Gibson and Brown 2009). For example, public policy produced by government departments is generally reliable as a source, providing a 67

88 degree of balance. In contrast, web pages or submissions from industry groups or NGOs have less reliability, but still provide an important source for analysing the policy process. The majority of primary documents obtained for the case study related directly to REDD+ and to the associated policy areas. These were publicly available on-line or from relevant agencies/organisations. Some key government documents in Indonesia are published with an English language version, however, media materials and other information are often available only in Bahasa Indonesian. Although the researcher has basic Bahasa Indonesia skills, translation assistance was required for a number of documents. Table 3.3 Document Sources Source Primary material Secondary material Government Policies, strategies, Reports, newsletters (national, provincial, district) legislation, grant applications, meeting summaries, websites, legal proceedings, maps Multilateral and Policies, reports, program Newsletters bilateral agencies and project documents, meeting summaries, contracts, websites NGOs and CSOs Policies, submissions, project documents, media Reports, newsletters, theses releases, websites Industry Submissions, project documents Articles Consultants/Research Technical and policy reports, websites, maps Working papers, background reports, articles, books etc. Media Photos Newspaper articles 68

89 Reliability and validity of data sources is important in case study design. One of the strengths of single case studies is that a combination of data collection techniques can be used (Yin 2009). Combining different sources is often referred to as data triangulation and provides one method for clarifying meaning and verifying interpretations (Stake 2005, p.454; Seale 2012, p.535). Triangulation with different data points can lead to corroboration or convergence around a particular social event or process, reducing bias and managing the validity of the analysis (Gibson and Brown 2009, p.70). Overall the researcher collected a relatively detailed and comprehensive body of data. To improve reliability and validity the researcher used triangulation methods to combine multiple sources of documentary data with interview transcripts. The complementary data offered greater depth and a more detailed and critical understanding, rather than only confirming analytical insights. Data triangulation also served as a method to compare and corroborate the perceptions of interviewees across the organisations. On some key issues, however, there was contradictory or non-corroboration across data sources and competing interpretations of the policy process. On these issues, care was taken to explicitly outline differences in opinion and understanding Ethical Conduct The research needed to address a number of ethical issues that can arise with data collection and the potential to expose individuals or groups to personal and professional risks (Tracy 2013). In a public policy context, researchers may seek comment on government policy or other sensitive and confidential issues, and compromise the interviewee. Often there is an uneven relation between the researcher and the community or organisation under study, and the potential for coercion of respondents who may be unwilling participants (Cannella and Lincoln 2011, p.82). The conduct of research in some locations and with some particular groups may also conflict with existing laws, or with religious and cultural norms of behaviour. These ethical issues applied to the research, as REDD+ has a high public profile in Indonesia and is politically sensitive. Research therefore needed to be sensitive to community interests around forests and particularly indigenous peoples, in addition to government agency staff engaged in program design and implementation. The key areas for ethical conduct in this research project focused on the fieldwork interviews, and particularly informed consent from participants, confidentiality, data security and issues of trust and accountability in the data collection process (Ali and 69

90 Kelly 2012). The research also needed to emphasise the voluntary nature of the research and ensure the accounts provided by interviewees were balanced and representative. In order to conduct interviews the researcher applied for ethical clearance from Griffith University Human Research Ethics Committee, and received approval under the Protocol GU Ref No: PPP/05/11/HREC. The research also complies with the Australian Government s National Statement on Ethical Conduct in Research Involving Humans. Copies of all ethics materials can be found in Appendices B and C. The researcher carried out the following steps to manage ethics in the interview process: A standard letter or was sent to potential agencies or organisations to explain the research project and requesting approval to conduct interviews with staff; Once approval was granted a research outline was sent to potential interview participants, who could then voluntarily contact the research team; Prior to each interview the interviewer outlined the risks and confidentiality issues with participants and then obtained informed consent, either written or verbally; To manage confidentiality all interview transcripts were de-identified, so that names and titles or interviewees were removed and replaced with a code key and pseudonym; and All interview recordings and transcripts are stored on protected files and are not publicly available. To conduct fieldwork, the researcher obtained a Foreign Research Permit from the Ministry of Research and Technology (RISTEK), and a temporary resident visa from Ministry of Immigration in Indonesia. This is a legal requirement for all research in Indonesia. The GoI visa rules required that the research be conducted in an ethical manner so as not to impact on Indonesian citizens. All government agencies with responsibility for REDD+ approved the project design, including the interview questions and participating organisations. The project was conducted in accordance with the strict requirements of the visa and permit. The Indonesian research partner was the Parahyangan Centre for International Studies, Department of International Relations, Parahyangan University, Bandung, Indonesia. The researcher spent 6 months in Indonesia between January and June 2012, to conduct fieldwork and as a visiting PhD student at Parahyangan University. The purpose and reasons for the extended stay 70

91 (beyond the 6 week data collection period) was that the researcher was a recipient of a Prime Minister s Australia Asia Award in 2012, issued by the Australian Government. The Award required a minimum 6 months stay in Indonesia, and involved other activities including lecturing and seminar presentations at the host university Data Analysis To organise and analyse the data the researcher used a combination of coding and thematic analysis. Coding provides a systematic method to collate and organise the data according to related topics, concepts and themes, and to link the text to the theoretical framework (Gibbs 2007; Saldana 2013). Codes can be developed from existing literature to identify the main variables and theoretical concepts. Codes can also be derived from raw data, such as interviews or documents, where close reading and synthesis leads to new topics (Rubin and Rubin 2005, pp ; Gibbs 2007, p.45). Thematic analysis involves aggregating and connecting the themes that emerge from the data according to similarities, relationships and any differences (Richie and Spencer 2002; Gibbs 2007). The researcher developed several coding and thematic frames during the data analysis and writing stages. The first stage involved the preparation of a descriptive coding frame based on topics identified in the REDD+ literature and from primary materials. The researcher used a combination of manual coding and the computer program NVivo to code all the interview data and documents. In the second phase, a coding frame was developed from theoretical framework (outlined in Chapter 4) and then used to re-code the data. The researcher also conducted thematic comparisons to identify the most relevant information and to establish linkages between the different units of analysis. The two coding frames, descriptive and thematic/theory based can be found in Appendix D. At several stages, the researcher also applied process tracing methods to understand the steps in the development of REDD+ policies, and the reasons, incentives and motivation behind several of the major bureaucratic reforms. Process tracing is essentially a method used to explain how mechanisms operate within a case (George and Bennet 2005, Parson 2010). The researcher used both deductive and inductive reasoning to evaluate the empirical material. Deduction offers as an interpretive lens to view the intent or purpose of different practices or phenomena using existing theoretical principles. Analytic induction allows themes to emerge from the case study data in order to develop new 71

92 theoretical concepts or generalisations (Moses and Knutsen 2007, p.136). During the data analysis stage, theory provided the deductive framework to interpret the complexity of REDD+ processes, the change and formation of new institutions, and the role and agency of multiple governing interests. The application of theory was critical to the identification of relationships across the large body of data, and to guide the analysis and writing phase. Importantly, the analysis aimed to create a dialogue between the theory and empirical evidence in order to generate a rich and theoretically informed narrative that explores the complex processes underway (Hay 2002, p.47). The detailed narrative presented in Chapters 5, 6, 7 and 8 aims to capture the interaction and relationships between actors, ideas and institutions in order to uncover and interpret the processes of change and reform in Indonesia. In addition, the empirical material on Indonesia was analysed inductively to expand and develop several theoretical concepts. 3.4 Conclusion Chapter 3 has presented and justified the methodology applied to this thesis. The research adopts an anti-foundationalist ontology, constructivism/interpretivism as the epistemology, and qualitative data collection and analysis methods. This approach was particularly suited to the research as it aims to provide a rich and theoretically informed analysis of REDD+. Further, the research aims to understand and interpret the reform process through the views and interests of the multiple actors shaping and participating in the different programs and activities. To collect data for the project, the researcher conducted semi-structured interviews with a cross section of representatives from government and non-government organisations, and these interviews were triangulated with primary and secondary documents. The Chapter argued that a single embedded case study provided a suitable methodology for the research. A single embedded case allows the major components of Indonesia s REDD+ program to be studied in depth, within the same context and across political jurisdictions from the national to the sub-national scale. Indonesia was selected due to the scope of the government s reform process, enabling detailed analysis of multiple programs and activities. Although a single case limits the broader application of the research findings, it enables exploration of the complexity and relationships between the multiple variables and components. The case study on Indonesia is significant because it provides one of the first critical applications of 72

93 governmentality theory to REDD+, and importantly the combined theoretical framework developed in this thesis. The next Chapter will introduce and outline this theoretical framework. 73

94 74

95 Chapter 4: Governmentality and Environmental Policy 4.0 Introduction Chapter 4 outlines and presents the theoretical framework applied to the case study on Indonesia. As discussed briefly in Chapters 1 and 3, the research utilises governmentality theory to analyse the development and partial implementation of the program Reducing Emissions from Deforestation and Forest Degradation (REDD+) in Indonesia. Governmentality theory is particularly suited to analysing the multiple activities and practices of governing, and to explore how power functions in contemporary government (Lemke 2002; Foucault 2007, 2008; Miller and Rose 2008; Dean 2010; Bröckling et al. 2010; Walters 2012). The theory focuses on how actors govern according to certain rationalities or thinking, and the implementation of government programs through associated technologies. Further, the theory investigates the role of different subjects employed in forms of individual and collective governing, and the function of knowledge systems and discourses in constructing domains of government intervention. Governmentality theory is widely applied to the analysis of government programs in liberal democracies, and increasingly to environmental policies such as climate change. To a lesser extent governmentality is used to analyse forms of democratic and authoritarian government in developing countries. There are, however, limitations with the theory. In particular, the application of the broader governmentalities of liberalism and neo-liberalism may be inadequate for investigating the various micro-processes and practices of governing that manifest in different political contexts. This Chapter thus argues that further development of governmentality theory is required in order to analyse the complement of programs and activities that constitute REDD+, how they are designed and how they operate within Indonesia. The Chapter outlines a theoretical framework that combines governmentality theory with three theories of environmental politics namely administrative rationalism, neo-liberal conservation and deliberative democracy (Dryzek 2005a, 2013; Kronsell and Bäckstrand 2010; Fletcher 2010). This combined theoretical framework will be used to analyse the major components to REDD+ including; regulatory, technical and bureaucratic reforms; the design of carbon markets and incentives; and the application of participation and engagement processes. The 75

96 purpose of combining governmentality with three environmental theories is to develop a framework that offers an expanded and more nuanced set of analytical devices to better understand the operation of REDD+ in Indonesia. The Chapter is organised into 5 sections. The first section provides an introduction to Foucault s theory of governmentality and power, and states why the thesis adopts a broad and flexible approach to governmentality. The chapter also convers the genealogical transition from sovereignty to bio-politics, and the operation of liberal and contemporary neo-liberal government. The Chapter then outlines the analytical components of governmentality that will be applied to this research, followed by discussion of the key criticisms and limitations. This follows with a review of the application of governmentality to green politics, and to climate change policy. The final section introduces the three environmental theories that will be combined with governmentality to form an analytical framework. Each of the three theories will be further developed in Chapters 5, 6, 7 and 8 then applied to the empirical case study material. 4.1 Governmentality Foucault (1991a, 2007, 2008) introduced the concept of governmentality in a series of lectures presented at the Collège de France in the late 1970s and early 1980s. However, Foucault s theories on governmentality were under development, and the concept subsequently shifted as the lectures progressed leading to a degree of ambiguity in its application (Walters 2012). What follows is a short overview of the various concepts of governmentality theory, and how these concepts will be applied to this thesis. The expansive literature on governmentality indicates several broad interpretations and applications of governmentality. For Foucault (1991), governmentality initially offered a concept to outline the emergence of government as a complex arrangement of governing practices, and its relationship to sovereignty and discipline as forms of power. In this context, governmentality provides a genealogical analysis of the emergence of government as a particular form of rule, and how certain practices evolved in the context of Western liberalism and later neo-liberalism (Lemke 2002; Walters 2012, p.27). Governmentality therefore explores new forms of thinking related to the exercise power in certain liberal societies, and principally new realities based on 76

97 economic management, and the focus on populations as the subject of government (Foucault 2007, pp ). Further, Foucault (2007, 2008) sought to elaborate how the population was managed through the apparatus of security, expanding from military and policing forces to include numerous institutions that support the functioning of society and the economy, such as schools and medical facilities. Foucault (1991a, p.102) argued in the context of early liberalism that governmentality comprises: The ensemble formed by the institutions, procedures, analyses and reflections, the calculations and tactics that allow the exercise of this very specific albeit complex form of power, which has its target the population, as its principle form of knowledge the economy and its essential technical means apparatuses of security. Whilst the early lectures relate to a more precise, historical based account of the emergence of liberal government, the later lectures offer a much broader and abstract application of governmentality (Walters 2012). Foucault s (2007) intention was not to outline discrete systems of power, but instead provide a broad approach to the analysis of different governing practices. This theoretical and analytical application combines two aspects of governing, namely the rational understanding or thinking about a phenomenon and acting upon the same phenomenon in order to change it (Miller and Rose 2008, p.15). In this application, governmentality or governmentalities refer to how systematic ways of thinking about governing are embedded within the practices of governing (Dean 2010, p.26). Governmentality is thus a form of thinking and reasoning that enables both those who govern and the governed to understand the nature as well as the various strategies or techniques of government. The concept of governmentality thus enables systematic scrutiny of the close relationship between techniques of power and knowledge, since thought is intimately embedded within governmental practices that are designed to shape and reform social conduct (Bröckling et al. 2010, p.2). The linkages between thinking about and seeking to enact government conceives of who or what are to be governed within society, and how certain social and economic processes are amenable to forms of calculation and planning (Rose et al. 2006, p.98). Foucault was thus drawing attention to particular kinds of political reasoning, to the emergence of particular knowledge and language, and to techniques for exercising rule and ultimately governing the state (Walters 2012, p.38). Governmentality thus offers an analytical approach to analyse the multiple forms of governing and the exercise of power, and importantly how governing techniques and 77

98 practices are subject to constant change. Governmentality as an analytical approach is not limited to rigid models and holistic logics of self government that characterise liberalism and neo-liberalism, but can encompass various programs of government, and multiple forms of rule (Rose et al. 2006; Collier 2009; Dean 2010). Indeed, there are multiple and heterogeneous approaches to thinking about government that involve numerous forms of agency and authority (Dean 2010, p.27). As Walters (2012, p.35) argues governmentality can encompass the particular practice or experience of governance and grasp it as a hybrid, a combination of different techniques, knowledges and rationalities. Rose et al. (2006, p.98) suggests that governmentality can be expanded to investigate how projects of political rationalisation are constantly undergoing modification in the face of some newly identified problem or solution, while retaining certain styles of thought and technological preferences. Governmentality should therefore be understood as a very flexible category that encompasses different styles of thought, their conditions of formation, the principles and knowledges that they borrow from and generate, the practices that they consist of, how they are carried out, their contestations and alliances with other arts of governing (Rose et al. 2006, p.84). Collier (2009) adopts a complementary approach arguing for a topological analysis that evaluates the redeployment, recombination, and patterns or correlation in governing practices and the application of power; and importantly to explore the governmental ensembles that emerge from these applications. For example, a topological approach can show how styles of analysis, techniques and forms of reasoning associated with advanced liberal government are being recombined with other governing approaches (Collier 2009, p.97). This thesis adopts and applies a broader theoretical orientation to governmentality that recognises the multiple and heterogeneous practices and approaches to governing, and importantly how the programs of government are subject to change and modification when applied in different contexts. As such the thesis recognises that contemporary forms of government deploy a whole series of intersecting governmental approaches, including (but not limited to) liberalism, neo-liberalism/advanced liberalism, bio-politics, security, sovereignty, and discipline (Foucault 2007, Dean 2010, Fletcher 2010). Critically, these overarching governmentalities are relatively flexible, yet comprise certain bodies of thought and knowledge, and employ combinations of practices and techniques that share similar traits or orientations. Further, the thesis recognises that the constant redeployment and recombination of governing approaches in different contexts leads to new variations or a hybridisation of governmentalities (Foucault 2007; Collier 2009). As such there is not a linear transition 78

99 from one governmentality to the next, but constant transformation in thinking and application about how rule is to be conducted. This broad and flexible application suits the orientation of this thesis and the case study that is applying governmentality concepts to the introduction of an international climate program in a non-western developing country. This thesis will explore how these broader governmental categories identified the literature are deployed in the case of Indonesia; how they encompass and combine more specific practices and strategies of rule; and how they change or are reconfigured in the context of governing forests and carbon. This chapter now introduces the theory related to specific governmentalities that will be applied in this thesis, a discussion that will be elaborated and expanded in chapters 5 through to Government and Power Foucault (1991a, p.102) re-defined government through the term the conduct of conduct. Government is not confined to public administration or the rule of law but represents a broader and calculated steering of human conduct. Government incorporates multiple practices that seek to direct or mould the behaviour of individuals to achieve certain outcomes, such as through regulating the economy or public education, but also in forms of self-regulation over the individual body. Foucault (2007) broadly referred to these practices or techniques as the arts of government. However, this application of government pre-supposes a certain level of freedom on individuals in their capacity to act (Lemke 2002; Dean 2010). Government operates by shaping these freedoms and moulding the capacity of individuals and collectives to regulate their own activities through combinations of economic, social and political intervention. In addition, rule by government is not exclusive to authority exercised through the administrative state, but rather multiple agents and authorities govern at different sites (Rose et al. 2006, p.85). Dean (2010, p.18) argues for the following expansive definition: Government is any more or less calculated and rational activity, undertaken by a multiplicity of authorities and agencies, employing a variety of techniques and forms of knowledge, that seeks to shape conduct by working through our desires, aspirations, interests and beliefs, for definite but shifting ends and with a diverse set of relatively unpredictable consequences, effects and outcomes. 79

100 The function of government is closely correlated with the operation or exercise of power. Foucault initially argued that political power exercised through government is distinct from sovereign and disciple as models of power formerly exercised by the crown (Foucault 1980, 2007). Sovereignty represented a more explicit and overt form of power exercised by an individual or centralised institution over the citizenry and over territory. The sovereign model of power was encapsulated in monarchical rule, and applied through juridical law and the suppression or restriction of activity. Repressive power or discipline, by contrast, attempts to control citizen s desires through restrictions on individual activity. Institutions such as schools, military training and religious instruction create particular regimes of discipline and certain sanctions that control and regulate human behaviour. In later work, Foucault (2008) outlines that government retains and utilises the institutions of sovereignty and discipline but reorganises and recodes them to appropriate different functions (Dean 2010, p. 29). For example, more modified forms of sovereign power function within representative democracy, operating through less oppressive means such as social laws and norms (Brown 2009). This approach then focuses on how sovereignty, discipline and government are recombined and reconfigured in complex arrangements or systems of correlation (Foucault 2007, p.8). By contrast Foucault (1977, p.9) argued that in liberal and neo-liberal government, power is more productive and is exercised at the micro level through a whole range of social relations, rather than negative forms of repression or overt authority from above. In advanced liberal government, for example, power is produced through networks that create and support self-regulating individuals and selfidentities, as much as controlling their activities through regulatory means (Miller and Rose 2008) Bio-politics, Liberalism and Neo-liberalism The lecture series presented by Foucault and resultant literature outline the development of government as a form of rule, and the thinking, practices and tactics that characterise how the social and the political is to be governed. The emergence of government as a form and expression of political authority occurred from the 18 th century onwards and was represented by a series of new tactics or practices designed to govern the population. Authorities gave increasing attention to managing and promoting the welfare of the population, to wealth creation, personal security and the improvement in health conditions and education (Smart 2002; Foucault 2008; Dean 80

101 2010). Termed bio-politics or bio-power these new practices aimed to exercise rule across the body of the citizenry and through the management of the individual (Nadesan 2008; Lemke 2011). Bio-politics encapsulates the processes by which government created certain methods for categorising and demarcating the population and the economy into problems that required intervention. Bio-politics is principally exercised through the sciences and statistical analysis and the collection of detailed information that enables aspects about people to be recorded and managed. Having a large body of data on the population or a particular aspect of the economy enables the administrative state to pursue a particular form of social and economic intervention (Lemke 2011, p.56). The bio-political techniques employed by the administrative state focus on management through policies, planning and strategies rather than pernicious forms of overt or repressive power. Governing the population through bio-politics coincides with the emergence of early liberalism. The aim of liberalism was to reconstitute and reconfigure the practice of ruling from the authority of the sovereign to government conducted primarily through the economy (Dean 2010, p.134). Liberalism was based on new theories of political economy that imposed certain limits or conditions on government action. Government was increasingly measured according to perceived truths about the importance of the economy, and these truths determined the purpose, character and effectiveness of different interventions. The focus and application of government policy was economic management, rather than the exercise of law and judicial power over the population. Government power was further limited by the increased focus on human subjects and a replacement of external regulation over the populace to governing through prescribed freedoms or liberties of the individual (Bröckling et al. 2010, p.26). These liberal freedoms comprise a calculated yet indirect form of exercising power, where individual action is not restrained but structured according to certain fields of possibility (Burchell 1996, p.26). Liberalism functions through the exercise of these individual freedoms or liberties that condition how society is to be governed through the implementation of various laws and norms. Thus government must operate according to a type of natural order that fosters the capacity of an independent civil society to generate its own wealth and wellbeing (Gordon 1991). The transition to liberal rule also witnessed the development of what Foucault (1991a, p.103) termed the apparatus of security whose purpose was to ensure that freedoms or liberties were protected. Security protected the rights of individuals to engage in the market by ensuring that social activities and groups were 81

102 regulated, but this regulation also provided a type of controlled liberty that managed any threats to a functioning economy and society (Dean 2010). The challenge or problem for liberal government is to balance the limits of individual action against the proper functioning of the market and civil society. Liberalism underwent further reconfiguration during the course of the 20 th century, first though a period of welfare liberalism followed by the rise of contemporary neoliberalism, and the closely associated advanced liberal government. As Miller and Rose (2008, pp.79-81) argue, forms of welfare, neo-liberal or advanced liberalism comprise governmentalities of rule that operate according to certain logics of intervention, and through certain practices or strategies of government. In welfare liberalism, the administrative state intervened directly in the economy to ensure the stability of society and to guarantee certain social benefits, such as employment and health care (Miller and Rose 2008, p.80). Neo-liberalism, by contrast, was a reaction the logic of the welfare state and involves a systematic and comprehensive expansion of the market - from a separate economic entity governed by specific laws to a system of relations governing everyday life (Gordon 1991, p.43; Foucault 2008, p.248). Rather than government structuring the economy to provide individual and collective goods, the market itself reconfigures how government operates in order to optimise individual benefits. This requires a downscaling of government intervention in society, and a restructuring of public administration and policy according to standards of market performance, competition and efficicncy (Bevir 2011, p.463). Accordingly, government becomes reliant on new types of expertise from economics, accounting and financial management and outsources responsibilities to the private sector or civil society (Mackinnon 2000; Miller and Rose 2008). The aim of neo-liberalism is to create conditions or incentives that produce competitiveness and entrepreneurial behaviour, and importantly to drive individual desire for self-benefit (Burchell 1996, p.23). 11 Government strategies focus on creating a social order of individual subjects who govern themselves according to rational calculation based on competitive self-interest and norms of personal responsibility (Rose 1996). Miller and Rose (2008) identify a further shift towards advanced liberal government; a contemporary political rationality that relates to a certain kind of problem-making, as well as to a characteristic set of techniques and styles of 11 Dean (2010) makes a distinction between neo-liberalism as an economic rationality of government, and advanced liberalism as the broader collection of rationalities, technologies and subjectivities that characterise the practices of governing in advanced liberal democracies. 82

103 reasoning and governing. 12 In advanced liberalism, government continues to operate through the freedoms and responsible agency of the individual, but without reference to a broader society (Dean 2010, p.194). Government practices aim to regulate, shape, structure and steer the individual through a plurality of practices, such as new forms of professional association and standards, types of personalised education and training, and through membership of community associations and networks (Miller and Rose 2008, p.213) State and Non-state Agents In governmentality analysis, the role and function of the state is reconfigured. In many historical and theoretical accounts the state is created as a specific and often fixed entity with a centralised political apparatus that governs over citizens, territory and resources. In contrast, Foucault (1991a, p.92) argued that the entity of the fixed or hegemonic state was a mythicised abstraction, and rather it represented a combination of governing functions, practices, tactics and calculations. This reformulation aims to understand how the modern administrative state emerged as a response to the particular problems of government (Lemke 2007; Walters 2012). In the shift from sovereignty to liberal rule the state underwent a process of governmentalisation and was reconfigured according to external demands and adopted new administrative authority (Dean 2010). Importantly, the state was institutionalised with new fiscal, organisational, judicial and security responsibilities in order to manage the population. The state also emerged as a particular type of government formation that positioned itself alongside and together with other institutions in civil society (Saar 2010; Jessop 2010). Miller and Rose (2008, pp.56-57) argue that in the contemporary context, the state can be seen to operate through these complex networks of relations found within an array of public and private organisations. The state thus represents a certain way of governing that responds to changing problems and emergent risks, and is itself subject to crises and transformation (Saar 2010, p.37). Governmentality analysis is therefore focussed empirically on the techniques and activities of the state as part of the wider strategies of government, and the relationships between the state and other political and non-political interests. 12 For the purposes of this thesis neo-liberalism and advanced liberalism will be applied as two separate but closely related governmental categories. Neo-liberalism relates more specifically to a shift towards markets as the governing principle in society; whereas advanced liberalism includes an additional shift in thinking, techniques and strategies for governing. 83

104 The role and agency of non-state actors is a problematic theme in governmentality analysis. Governmentality theory recognises that political authority in liberal and neoliberal government represents a plural form of agency involving multiple actors (Rose et al. 2006, pp.97-98). Whilst the modern state employs many of the functions of government, it shares and facilitates governing through multiple authorities and interests, whether from civil society or the private sphere, collective or individual. Sending and Neumann (2006, p.652) contend that the increased function of these nonstate actors in mobilising the preferences of the population and carrying-out regulatory functions is representative of the changing rationalities of neo-liberal government. Non-state interests are therefore redefined from an object of government action to an active participant in, and subject of, governance. By becoming participants and subjects of rule, non-state actors are taking part in these relations of power and are therefore reproducing and transforming its operation (Bryant 2002, pp ). This is because the self-association and political will formation characteristic of civil society and non-state actors do not stand in opposition to the political power of the state, but is a most central feature of how power operates in late modern society (Sending and Neumann 2005, p.657). Indeed, non-government organisations (NGOs) working on community development programs have often displayed an increased capacity to deliver services in places where previous government run programmes have failed. 4.2 Analytical Framework The discussion above outlined the main concepts and tenets of governmentality theory. This next section extends the decision from above by outlining the key elements of the analytical framework that will be used to investigate the various policy and program components adopted for REDD+ in Indonesia. Governmentality theory offers a relatively well developed and widely applied framework to research the multiple and often intermeshed practices and programs of government. Governmentality analysis focuses on the conditions or circumstances under which different regimes of practices and associated programs of government are developed and implemented, and how they are maintained and then reformed (Dean 2010, pp.30-31). The framework developed in this section integrates the key concepts used in governmentality analysis, including; a defined rationality and associated technologies; the definition of certain 84

105 problems (or problematisation); the application of bodies of knowledge and discourse; and the formation and agency of subjects This framework is drawn from several widely used approaches to governmentality analysis including Lemke (2002), Miller and Rose (2008), and Dean (2010). Figure 4.1 outlines the key governmentality concepts. Rationalities Subjectivities Technologies Governemtality Discourses Problematisation Knowledge Figure 4.1 Governmentality Concepts (Adapted from Lemke 2002; Miller and Rose 2008; Dean 2010, Bröckling et al. 2010) Rationalities and Technologies The central concept in governmentality analyses focuses on the particular mentalities or rationalities that underpin governing practices. As discussed above, rational thought provides a method to conceive of reality so that it is amenable to particular methods of calculation and reasoning, often using accepted or formal bodies of knowledge or expertise (Gordon 1991; Miller and Rose 2008; Olsen 2008; Dean 2010; Bröckling et al. 2010). Rationalities therefore relate to systematic ways of thinking about how to conduct rule and how society is actively governed (Gordon 1991; Lemke 2002; Miller and Rose 2008). They comprise a collective logic around the confusing activities of government, and thereby sort the multiple legal, regulatory and bureaucratic processes 85

106 into logical associations (Olson 2008). Rationalities also produce ideals for how government should be conducted, such as promoting economic growth and political freedoms, and they articulate the domain of intervention into such areas as the economy, population, or over geographical territory. It follows that rationalities collectively define the goals, rules and procedures for action and thereby define how government is to achieve certain outcomes (Miller and Rose 2008). As such, there exist multiple and often competing rationalities that are regularly subject to challenge or failure, often leading to crises and reform in governmental practice. Rationalities are closely associated with different technologies or techniques of governing. Whilst rationalities conceive of certain ways of thinking about governing, technologies enable this thinking to be operationalised and applied in action (Miller and Rose 2008; Lövbrand et al. 2009b). Technologies represent the specific tools or mechanisms for governing and therefore enable authorities to shape, direct and limit collective and individual activities. Technologies are an expansive collection of various mechanisms, devices, procedures, policies, techniques of calculation, processes of analysis, and materials through which authorities have sought to control or steer the behaviour of populations (Miller and Rose 2008). They are also subject to rational thought and deliberation and are often designed for specific purposes. At any one time there are multiple technologies in operation, often linked to diverse rationalities and the practices of modern government and the state (Lemke 2007). In this regard, technologies are not limited to a function of particular rationalities but also come to define the domain of intervention. Thus the application of technologies forms an important component in the operation and exercise of power. Governmentality analysis should therefore focus attention on the production of particular technologies, how they are deployed and their relationship to particular forms of authority (Miller and Rose 2008, p.37). Although indispensable to governing, technologies remain a relatively under-developed concept in the governmentality literature. There are, however, several categories of technologies that operate within liberal and advanced liberal government, including technologies of the self, citizenship, performance and agency (Sending and Neumann 2006, p.657; Dean 2010, pp ). Technologies of the self are mechanisms or practices that enable individuals to pursue guidance and regulation over personal affairs, creating enterprising, self-managing and competitive citizens (Burchell 1996). These mechanisms operate though the market to create moral and conscious consumers, but also as norms and techniques of self-improvement such as health 86

107 regimes and lifestyle choices (Lemke 2002; Rutherford 2007). Technologies of performance are designed to build trust and professional accountability, such as through audits, codes and systems of accreditation and other practices aimed at keeping subjects performing to certain standards (Dean 2010, p.196). Technologies of agency or citizenship relate to processes for expanding the opportunities for individuals to engage in the practice of government. Examples include processes of engagement, partnership and negotiation that seek to empower citizens or social groups, but also through contracts for delivering certain government functions (Dean 2010, p.196) Problematisation, Knowledge and Subjectivities In addition to technologies, governmentality theory focuses on the construction and definition of problems. Government is essentially a problematising activity in that those who govern constantly create or re-identify problems in order to call into question the current regimes of practices (Miller and Rose 2008). Most political issues at any scale are framed as problems, for example poverty and unemployment, or the need to recognise the rights of minorities. Here the existing regimes of governing and the particular institutions of government fail to serve the purpose for which they were designed (Dean 2010). Government action is therefore routinely tied to how complex problems are constructed and made visible. For example, liberal government and the evolution of the modern state represent particular responses to complex problems that emerged in relation to management of the economy, the population and the welfare of the people (Foucault 2007). The prescription and application of certain programs of government are therefore closely aligned with the way in which different agents construct problems and define social reality (Bröckling et al. 2010). Knowledge and expertise are also central features in the analysis of the different rationalities and technologies of government. Rationalities and ways of thinking about government are contingent upon, and are embedded within, particular forms of knowledge, such as ideas, philosophies, theories and language (Foucault 1980; Miller and Rose 2008; Dean 2010). In the modern liberal economy, information about economic conditions, and the language and ideology of the market comes to shape how we think about the role of government intervention, and broader social relations (Rose et al. 2006). Thought and knowledge are therefore closely tied to how power functions within the many techniques and actions of governing (Foucault 1980). The application of knowledge by different agents plays a fundamental role in creating realities about the natural and social world so it can be subjected to analysis and 87

108 calculation (Miller and Rose 2008). Creating knowledge about the population, the individual or the environment enables governing authorities to observe and define the fields of intervention. Policies or programs adopted by the state or by other agents are thus often created and legitimised through different forms of knowledge (Lemke 2002). In the environment field for example, science offers an epistemological process to identify physical and social components of the earth, and to legitimise certain practices and instruments designed to regulate or shape these components (Lövbrand et al. 2009b, p.8). Generating knowledge is also closely allied to which actors have authority or legitimacy to act and speak on certain issues. Knowledge is produced by particular actors and institutions who determine which type of information is legitimate and what lines of enquiry should be pursued (Rutherford 2007). A further two components to the governmentality framework are discourses and subjectivities. Foucault (1980, p.131) argued that political power and economic production demand certain truths about their effects, and these truths are propagated through discourse. Discourses express particular ideas, concepts and representations that ascribe meaning to social phenomena and importantly act to construct different realities (Hajer 1995; Hajer and Versteeg 2005). Discourse in this sense has a productive function as they construct the truth about subjects, but they also constitute subjects according to particular truths and understandings. In governmentality analysis, discourses are employed within certain rationalities, operating in conjunction with other technologies and practices of government, rather than constituting all realities and the subjects to be governed (Miller and Rose 2008, p.30). In a policy and programmatic context, agents often employ different discourses to advance particular agendas or certain values (Dryzek 2005a, p.14). Discourses also produce certain subjectivities around people and populations that shape the ideas and beliefs of actors and how they interact with society. These subjectivities represent the particular ways that actors create or construct individual or collective identity, which influences how they act or perform in the world (Bröckling et al. 2010). Subjectivities or identities also symbolise how actors respond to particular government programs, including their implicit support and at times resistance (Ball and Olmedo 2012). Dean (2010, pp.43-44) argues that identities are often shaped within and by the rationalities and technologies of government and through associated knowledge systems. In the neo-liberal context, for example, subjects are often cast as the self-conscious citizen whose identity is shaped as the consumer motivated by individual benefit and self improvement. 88

109 4.3 Critiques and Challenges The development of governmentality theory and the broader application to contemporary social and political issues has also generated a series of critiques and challenges. Criticisms are often made of Foucault s poststructuralist accounts of power and the relatively circumscribed or ambiguous role of the state. The concept of government power as productive and a function of social relations is often criticised for its totalising effect that precludes other models of power (McKee 2009; Bröckling et al. 2010; Walters 2012). The genealogical development of power in governmentality theory creates an almost monolithic concept that is embedded within liberal and neoliberal forms of rule (Bevir 2010). Power becomes a ubiquitous aspect of human interaction, expressed as rationalities that continually determine the field of human opportunities whilst limiting the potential for agency. In addition, governmentality studies often focus on agency as exercised from above and through those rationalities and discourses that condition the regimes of governing. This limits the opportunity to analyse resistance to power from below, and the contestation that manifests in many social and political settings (O Malley et al. 1997, p.510). This lack of attention to resistance neglects the ways that certain agents shape or are subsumed within particular practices of government, or indeed create their own particular rationalities and counter discourses to those prevailing in any given domain (McKee 2009, p.474). Critics have also argued that there is limited focus on difference or recognition of inequalities such as ethnicity, social status, location and gender (Rutherford 2007, p.301). A further critique relates to the dominant rationalities that shape government action. Governmentality theorists emphasise the study of specific practices rather than the exploration of ideal types of government (Lemke 2007; Dean 2010). However, critics argue that the rationalities and subjectivities of liberalism and neo-liberalism function as ideal types or as set of relatively inflexible master categories (O Malley et al. 1997). This rigid and often universal approach prevents the further rationalisation and transformation in thinking around governmentalities, or contributes to the development of new concepts and applications (Collier 2009; Bröckling et al. 2010; Death 2013). The homogenous nature of rationalities and technologies results in an inability to account for the constitutive hybridism of discursive patterns and mechanisms of power. Studies in governmentality can therefore fail to address the deviation or limits of government programs, or provide suitable accounts of alterative and hybridised forms of rule. Further, there is often limited correlation between the study of specific 89

110 mentalities of rule and the social or material conditions in which they occur (O Malley et al. 1997, p.512). In the case of neo-liberalism, governmentality approaches provide few coherent linkages between the individualised practices of rule and the contingent cultural processes that construct subjectivity and self-identity (Barnett et al. 2008). Many cases of governmentality studies are thus divorced from empirical reality, identifying discursive trends rather than attending to the confusing regimes of practices and the intersecting array of subjects who constitute processes of governing. Criticisms levelled at the theory relate to its application and theoretical orientation in non-western countries dealing with postcolonial legacies, and various forms of political authority. Critics such as Joseph (2009; 2010) and Chandler (2009) question whether governmentality has a universal application in the context of an international liberal or neo-liberal set of practices that govern people and populations in different jurisdictions. Governmentality is readily applicable to modern western democracies with suitable social conditions that give rise to certain techniques of neo-liberal government. However, these processes are not easily transportable to less developed countries that do not have the social relations in place to support neo-liberalism (Joseph 2009, p.426). In these cases it is difficult to apply governmentality theory to explain how populations are governed. For example developing states are often funded by international organisations to adopt neo-liberal democratic provisions of good governance defined as the capacity to implement and manage market reforms (Löwenheim 2008). Critically, governmentality theory has not adequately dealt with how some developing states introducing sweeping economic reforms yet failed to import democratic governance principles. There is also a tendency to assume that the expansion of capitalism utilises neo-liberal power strategies that seek to govern through individual freedoms and a competitive market (Joseph 2010). This ignores that capital expansion often employs disciplinary power in different contexts and that authorities in many countries do not readily govern according to liberal subjectivities but often employ violent and illiberal strategies. Whilst these criticisms are legitimate they also create further opportunities to apply governmentality theory to developing countries. Indeed, the orientation towards mentalities and technologies of power opens up a critical theoretical domain to analyse the multiplicity of governing approaches (Rose et al. 2006). This includes the operation of centralised state bureaucracies, plural forms of deliberative governance, market oriented network arrangements, and localised forms of ethnic or traditional authority. Several theorists (Dean 2002; 2010; Jeffreys and Sigley 2009; Hodge 2012) contend 90

111 that governmentality can account for authoritarianism and state administration in developing countries. Authoritarian approaches often combine liberalism s selfregulating society with sovereignty and disciplinary power employed within the apparatus of security (Dean 2010). In Fiji for example, quasi-authoritarian regimes support a range of economic freedoms, yet continue to limit political rights and often retain obedience and exercise coercive strategies to regulate the activities of civil society and ethnic minorities (Hodge 2012, p.1147). An aim of this thesis is to apply governmentality theory to Indonesia; a post authoritarian polity in transition to a liberal democracy, yet with a relatively centralised state administration combined with informal clientelist decision-making. This thesis will focus on the introduction of elements of democratic liberalism, neo-liberalism/advanced liberalism and bio-politics in the context of REDD+, and how these governing approaches adapt and contrast to existing patterns of authority in Indonesia. This thesis will therefore contribute to an expanding field of research that applies governmentality theory to policy reforms in different political regimes in non-western countries Governmentality Compared to Governance Chapter 2 outlined several limitations with governance theories and the need for alternative analytical frameworks. Governmentality theory is often contrasted, but also conflated with theories of governance, and therefore some clarity is required in the orientation of each theory. For the purposes of this thesis, governmentality and governance are considered distinct theoretical approaches that share some commonality but have some clear differences in analytical concepts and their application (Bevir 2011). There are affinities between governmentality and governance theories, particularly as they both address the role of public and private actors within the context of neo-liberal reforms. Governance and governmentality theories extend the scope of political analysis beyond the focus on institutional politics and are attendant to political power beyond the state (Lemke 2007). There is the shared concept that power is decentred 13 Notable research includes analysis of China s socialist administrative state and the rollout of market neo-liberalism (for example Jeffreys and Sigley 2009; Sigley 2004), the politics of community development in Asia (Watts 2003), and community development and environmental programs in India and Indonesia (Agrawal 2005: Corbridge et al. 2005; Li 2007). 91

112 and operates at the micro-level through networks within civil society, but also within formalised arrangements such as found in public private-partnerships. Both governmentality and governance theories recognise that under neo-liberal rule there has been a distinct shift away from the monolithic and centralised bureaucracies towards decentralised governing and the fragmentation of public agencies (Bevir 2011). Each theory also attends to the ascendancy of the market in reconfiguring how political and social programs are enacted, and their effects on government, society and the individual. Governance and governmentality theories have a number of critical differences. Governance theories still accept the duality of state and society, focussing on the hierarchical and often ineffective interventions by bureaucratic organisations. Governance compares the state to decentralised societal mechanisms, often investigating the interdependencies and networks operating between the two domains (Lemke 2007; Bevir 2011). Much of the governance literature focuses on alternative forms or modes of government that govern through indirect practices, in order to address an increasing complexity of political problems. These indirect approaches indicate governance is about steering society through the development of efficient instruments of problem management and policy making, often emphasizing stability (Pierre and Peters 2000; Enroth 2014). In contrast, the state in governmentality theory is an integral part of governmental practices, an instrument and effect of governing, and the tension with civil society represents a contingent part of governmental practices. Governmentality pays attention to the relationship between knowledge, technology and forms of intervention, and how these are brought into government strategies through discourse and at times political contestation and conflict (Lemke 2007). Indeed, contestation and political struggle are recognised as both elements and strategies of rule, and as such part of the problematisation process that catalyses reform. Governmentality therefore considers complexity and the modes of governance as particular style of rule or art of government adopted by the liberal state. In this context, governance can be viewed as a type of technology and discourse of liberal or neo-liberal government focused on stability, participation, consensus, coordination and collective problem solving (Rose et al. 2006; Lemke 2007; Bevir 2010). Governance thus suggests a form of governmentality promoted as the solution to problems of the (neo) liberal democratic state. 92

113 4.4 Environmental Governmentalities Governmentality theory is also widely applied to environmental politics and policy making, and there is a growing body of research on climate change mitigation and forestry issues. Green governmentality essentially applies Foucault and other theorists to unravel the overlapping governmentalities that manifest in green politics and environmental governance, focussing on neo-liberalism and bio-politics, but also sovereign and disciplinary modes of government power (Luke 1999; Rutherford 2007; Fletcher 2010). Green governmentality literature also investigates the ways in which the environment and nature is governed through the formation of certain subjectivities (Agrawal 2005). For example, neo-liberal green governmentalities can foster subjectivities of the virtuous and caring individual, who adopts practices to live sustainably, such as eco-purchasing and consumption practices (Rutherford 2007; Lovell et al. 2009). Through these technologies of the self, individuals become the subjects through which interventions are deployed, making citizens responsible and capable of environmental care (Rutherford 2007, p.298). Other environmental programs adopt Foucault s notion of discipline as an approach to governing, fostering environmental subjects in different communities by diffusing ethical norms of behaviour, often through public education and training, and aimed primarily to discourage deviance and immoral behaviour (Agrawal 2005, p.32; Fletcher 2010). Sovereign governmentalities, by contrast, can involve a state sanctioned protectionism over land and biodiversity, such as national parks that in many locations require active policing and enforcement (Fletcher 2010, p.177). The field of green governmentality also applies Foucault s insights on knowledge and discourse to environmental politics. According to this synthesis, scientific knowledge is fundamental to the creation of certain environmental truths that actively shape policy and institutional change (Luke 1999; Bäckstrand and Lövbrand 2006; Lövbrand et al. 2009b). In this assessment, ecology and the earth sciences are not independent of politics and rule. Rather, by constructing threats and risks to nature, science has become fundamental to the production of regimes of governmentality that recasts nature and the earth as a physical entity in need of expert management (Lövbrand et al. 2009b; pp.7-8). The requirements to sustain life are subsumed within the need to manage and harness the environment in order to protect humanity. The intense focus on understanding the life and wellbeing of the environment resonates with Foucault s notion of bio-politics, focussed on identifying, recording and modelling the statistical risks to nature, and then designing efforts to manage these threats (Darrier 1999; 93

114 Rutherford 2007). Natural sciences thus operate as a type of power/knowledge regime, where the power to produce knowledge about the environment is key to formulating the terms of its management (Rutherford 2007, p.298). Expert knowledge about the environment contributes to discourses regarding what constitutes appropriate government interventions. The expert scientific approach is displayed foremost in techno-managerial policies advanced under the stewardship of nature and sustainable development discourses (Hajer 1995; Luke 1999; Bäckstrand and Lövbrand 2006). Luke (1999, p.149) argues that discourses promoting the stewardship of nature often demand highly technocratic government, a process that reinforces the authority of the centralised administrative state. Sustainability discourse requires that current unsustainable practices need to be broken down and redesigned according to certain disciplinary norms of bureaucratic management. Discourse is also applied to the analysis of forest management in different countries, exploring how forest and conservation science plays a critical role in exerting power over people and resources (Winkel 2012). In many developing countries, contemporary forest policies often have their origins in colonial discourses that rationalised management in centralised agencies and promoted economic production from timber. More recent global discourses regarding sustainable forest management and environmental degradation are used to legitimise further state intervention in forest policy and into the lives of local communities (Asher and Ojeda 2009; Medina et al. 2009). These discourses apply science to determine what good forest management should be, whilst marginalising and discrediting other activities such as subsistence agriculture and cultural practices related to forests (Ambrose-Oji et al. 2002, p.159) Climate Change Governmentalities Climate change provides an emerging area of research on the application of governmentality theory (Stripple and Bulkeley 2013). Research suggests that from the outset the problem of climate change was framed by bio-politics and more specifically the global view created by science and earth systems modelling (Oels 2005; Bäckstrand and Lövbrand 2006). Bio-power is prevalent with the intervention of the state into all sectors of the economy for the purpose of monitoring and mitigating emissions (Oels 2005, p.198). The establishment of the The Intergovernmental Panel on Climate Change (IPCC) and the selection of scientists and other experts focussed efforts on developing complex emissions modelling and accounting frameworks. This modelling exposed the threat of rising greenhouse gas emissions and rationalised the 94

115 early techno-managerial policies of planetary wide mitigation and adaptation programs. The application of techno-scientific knowledge and methods of carbon accounting and standardisation has also made vastly different types of greenhouse gasses commensurable with one another (MacKenzie 2009; Lövbrand and Stripple 2011). These devices of greenhouse gas (GHG) measurement make a large field of natural and social systems comparable in terms of their stocks and flow of carbon and translate them into a global model (Lövbrand and Stripple 2011). Carbon accounting and the creation of carbon credit offsets operate as a regime of calculation, and thereby establish the carbon economy as a new domain of climate governance (Lövbrand and Stripple 2011, Methman 2013). In addition to bio-political analysis of climate change is the application of advanced and/or neo-liberal governmentality. The construction and transition to an advanced liberal governmentality of climate change involves a series of constitutive processes (Stripple and Bulkeley 2013, p.12). Whilst a supporting framework of technical science is maintained, the discourse and focus of policy interventions shifts to market based mechanisms at global and national scales (Oels 2005; Bäckstrand and Lövbrand 2006; Methmann 2010). The power to initiate climate interventions is now subject to market dynamics that supports dominant economic discourses focussed on paradigms of least cost abatement and green growth (Methmann 2010, p.367). This carbon economy is produced through the creation of carbon products and credits and regulated according to a series of everyday practices of market management (Lovell and Liverman 2010; Bumpus 2011). Markets also enable the state and other actors such as climate professionals to govern at a distance, such that regulating emissions is subject to third party standards and accounting processes (Methmann 2013, p.85). Neo-liberal climate change policies extend to a range of private practices that are variously administered through technologies of agency and performance. These include voluntary carbon neutrality programs and carbon auditing that benchmark and evaluate the relative value of individual or private emissions reduction programs (Paterson and Stripple 2012). Analyses of climate change policy and programs also investigate the construction of various subjectivities. From a neo-liberal perspective, Patterson and Stripple (2010) explore various voluntary carbon initiatives that re-cast private business and civil society as the active agents and subjects through which climate interventions are implemented. Carbon neutrality, foot-printing, dieting and rationing programs also function as a type of carbon conduct, a certain moulding and mobilising of subjectivities 95

116 around individuals and companies as carbon emitters with the responsibility to govern their own emissions through offsetting and purchasing practices (Patterson and Stripple 2010, p.359). These programs do not necessarily operate as international treaties enforcing rules over states, but act through subjects by steering their identities and normalising behavior (Patterson and Stripple 2010, p.360). According to Lövbrand and Stripple (2012, p.659) the increasing involvement of private operators and other actors in the carbon economy represents a transformation of politics that involves a replacement of formal and hierarchical techniques of government with more indirect regimes of economic calculation. Other approaches (Rutland and Aylett 2008; Rice 2010) pay attention to how local governments map out territorial climate spaces that establish certain permissible mitigation and adaptation activities to inform the conduct of local constituents. By engaging as subjects in adaptation planning, people and groups adopt the objectives of the state and are moulded into networks of selfregulating and responsible climate citizens (Rutland and Aylett 2008, p.640) Governmentality scholarship has recently shifted to REDD+, but this remains a developing area of research (see Gupta et al. 2012; Boer 2013; Stephan et al Methman et al. 2013; Lovell 2013). As discussed briefly in Chapter 2, REDD+ is subject to a series of practices that involve the quantification, standardisation, simplification and commensuration of GHG emissions. Commensuration requires the conversion of avoided deforestation into metric tonnes of carbon dioxide, and the ability to convincingly construct a business as usual baseline, or a counterfactual estimate of what would occur in the absence of the REDD+ measures (Stephen 2013, p.162). Commensuration also requires processes of marketisation, in order to express carbon projects in the language of the market and make them temporally and spatially exchangeable (Methman 2013, p.80). These calculative practices lead to the commodification of avoided deforestation, and combined with the monetisation process around markets lead to the carbonification of forest ecosystems (Stephan 2012, pp ). REDD+ measurement, monitoring, reporting, and verification (MRV) practices thus represent a type of bio-politics designed to make forests governable though the collection of data and framing of knowledge in a way that supports various policy interventions (Lovell 2013, p.193). However, MRV also operate as a type of duel accountability mechanism to measure a country s performance against international benchmarks and targets (Gupta et al. 2012). Other recent approaches (Stephan et al. 2013; Stephen 2013) have sought to apply a governmentality approach to REDD+ carbon markets. Stephen et al. (2013) combine 96

117 theories around hegemony with a governmentality approach on the micro-practices of power. The authors utilise the concept of articulation of different discursive elements to identify the bottom-up interaction between the governmentalities of bio-politics and advanced liberal government. They argue (Stephan et al. 2013, p.64) that it is not possible to identify one ideal type of governmentality informing practice, and that multiple technologies and linguistic practices are employed in REDD+. However, these contending discourses essentially reduce the visibility of multiple forestry and REDD+ processes to one of carbon stocks and tradeable markets credits (Stephen 2013). At the global level, REDD+ has thus morphed into the hegemonic discourse around forest carbon markets as a solution to climate change, one that has generated a broad level of support despite early resistance. The hegemonic process has depoliticised REDD+, shifting the debate from a discursive struggle around equity and justice to a technocratic discussion about carbon abatement and forest management (Stephan et al. 2013, p.73). 4.5 Developing an Expanded Theoretical Framework The discussion above suggests governmentality theory provides a suitable framework to analyse the ways in which the programs of government seek to govern people and carbon. However, there are several limitations or gaps in current applications of governmentality theory to REDD+ and climate change policy. Current approaches to REDD+ have focussed on the highly technical components of carbon measurement viewed through a bio-political frame, in addition to the broader governmentalities of advanced liberalism or neo-liberalism as applied to carbon markets. Whilst these applications provide a valuable framework there is a need to understand the complex, integrated and often confusing array of micro practices, forms of knowledge and subjectivities that operate to make REDD+ governable in different jurisdictions. Further, applying ideal and dominant types of governmentalities can lead to stable and uniform descriptions that mask the heterogeneity of practices, and make it difficult to account for the contingency and specificity of social relations (Stephan et al. 2013). To avoid over-emphasising and limiting the narrative to advanced liberalism requires using governmentality to study the highly varied practices and rationalities of climate programs found in different settings (Death 2013, p.82). Although REDD+ has an international dimension, its application occurs at the domestic level where a whole series of contingencies, social relations, localised politics and 97

118 existing forest governing regimes are in effect. Further, REDD+ involves a crosscutting array of policies, programs and initiatives that expand the rationalities and technologies of governing beyond markets and carbon accounting. Detailed analysis of these components therefore requires an expanded theoretical framework that enables an evaluation of the contrasting practices of governing, how they are deployed in particular countries or locations, and the agency and authority of various interests engaged in governing. This Chapter now turns to the development of a combined theoretical framework that can be applied to REDD+ in Indonesia. Governmentality theory can readily be combined with other social theories. Foucault s intention was to provide a toolbox on how people govern and are governed, and thereby opening up opportunities to combine governmentality with other compatible theories and analytical approaches (Rose et al. 2006; Foucault 2008; Stripple and Bulkeley 2013). Further, governmentality theory offers a broad analytical approach, one that adopts and integrates different social theories to interpret and challenge social relations and develop new ways of thinking about the practices and processes of governing (Bröckling et al. 2010). One promising development of green governmentality theory focuses on the rationalities of environmental governance. Kronsell and Bäckstrand (2010, p.30-32), for example, develop an introductory typology of environmental rationalities and combine this with governance forms or modes. The authors identify three broad rationalities that define environmental policy, namely administrative, economic and deliberative/democratic. These rationalities correspond with the governance modes of hierarchies, markets and networks, and relate to how people and institutions are organised and authority is exercised. This typology draws on Dryzek s (2005a, 2013) analyses of environmental discourses, and the major developments in environmental politics and policy in the last half-century. Dryzek (2005a, p.73) argues that these three discourses of administrative rationalism, economic rationalism and deliberative democracy correspond to the three main ways to coordinate environmental issues and represent distinct approaches to problem solving and policy formation. This thesis utilises the typological framework of environmental rationalities and discourses outlined by Kronsell and Bäckstrand (2010) and Dryzek (2005a, 2013) as a starting point. The next section provides a short overview of these government rationalities. 98

119 4.5.1 Administrative, Economic and Deliberative Approaches Administrative rationalism was the primary approach adopted by governments in the response to environmental problems in the 1970s, and remains fundamental to public management of environmental issues. In the environment domain, administrative rationalism prioritises the role of professionalised government agencies and the application of science to identify environmental problems and develop solutions (Dryzek 2005a, pp.76-84). The authority of these bureaucracies is often exercised through the production and control of technical knowledge, and the management of (often classified) information (Torgerson 2005, p.16). The broader mode or form that underpins the administrative state is one of hierarchy and the centralisation of decision making at the apex. Complex environmental issues are separated into component parts and often dealt with by different administrative sections or agencies. Policy tools employed by the administrative state include more traditional command and control type regulations and planning frameworks, and the application of environmental impact assessment. However, the centralised bureaucratic management of environmental issues has come under sustained criticism for its limited capacity to address multiscalar and multi-jurisdictional environmental problems (Torgerson 2005). Governments have struggled to apply regulatory standards and enforce laws, leading to implementation deficits and poor environmental performance (Dryzek 2005a, pp.92-93). Economic rationalism arose with the neo-liberal turn in government policy in the 1980s and an associated discourse that criticised state administration and regulation as a cause rather than a solution to economic and social problems. Theories of neoliberalism and associated economic rationalism argue that setting a price on ecosystem services provides the optimal and most cost effective solution to resolving environmental problems, because it changes the incentives that drive individual and collective behavior (Dryzek 2005a; Büscher et al. 2012). The policy mechanisms of choice include ecosystem markets, and pollution taxes and fees that charge companies for certain pollutants. While the focus is on individuals and companies, centralised bureaucracies are relegated to a secondary role of designing and operating the market instruments to ensure they operate efficiently. Critically, economic rationalism also faces some substantial challenges in responding to environmental problems. Although market and incentive mechanisms have been widely promoted they are problematic to implement across a range of environment and natural resource issues (Dryzek 2005a; Castree 2008). These problems have arisen, in part, because 99

120 markets do not deal with the complexity of ecosystem services, or account for nonmarketable environmental and social values (see Heynen et al. 2007; Bakker 2007; Fletcher and Breitling 2012). In contrast to both administrative and economic rationalities, deliberative approaches argue for interactive problem solving and the engagement of multiple stakeholders in decision-making (Baber and Bartlett 2005; Dryzek 2010). The premise of deliberative democracy is that rational individuals and groups can deliberate on environmental problems and reach informed and negotiated solutions (Lövbrand and Kahn 2010, pp.50-51). The underlying logic is that better environment decisions result from processes that are inclusive and participatory, leading to greater accountability and legitimacy of policies or programs (Dryzek 2010). Deliberative rationalism involves a number of participatory processes that attempt to include multiple citizens or interest groups in policy making, or in conflict resolution. Governments, for example, use consultation and public comment periods on new policies and legislation, in addition to more structured multi-stakeholder planning and citizen s juries (Smith 2003; Dryzek 2005a). However, the benefit of deliberative and participatory approaches remains ambiguous. Consultation processes can often be tokenistic and fail to address stakeholder agency, or meaningfully expand the democratic rights of participants, and particularly marginalised groups (Young 2003). In many cases engagement and participatory processes have not produced any substantive policy change or delivered improved environmental performance, either by government or by the broader community (Kronsell et al. 2010, p.220) Applying the Theoretical Framework As outlined in Chapter 1 this thesis will develop an expanded analytical framework by combining governmentality theory with these three theories drawn from environmental politics. Governmentality, therefore, provides the macro theory that is combined and applied to each of these three meso-level environmental theories - as outlined in Figure 4.2 below. The rationale is that integrating these theories offers a more comprehensive and nuanced set of analytical devices to study the multiple and contrasting programs of REDD+. In Chapters 5, 6 and 7 each environmental policy theory will be synthesised and analysed through the governmentality concepts outlined earlier in this Chapter covering; rationalities, technologies, problematisation, knowledge type, subjectivity, and discourses. Further, each Chapter will integrate these environmental theories with the broader governmentally theory outlined in section

121 above including concepts around bio-politics, liberalism, welfare liberalism, neoliberalism, advanced liberalism, discipline and sovereignty. This approach recognises that these broader governmentalities, whilst adopting certain styles of thought and orientation, are highly flexible and often accommodate different approaches or techniques to governing (Rose et al. 2006). Importantly, they are constantly subject to change and recombination when applied to different settings. This combined theoretical framework is then used to analyse the major components to REDD+ in Indonesia. In Chapter 5 the thesis will integrate governmentality theory on liberalism and biopolitics with theories of administrative rationalism, and then apply this to the introduction of regulations and reforms to the bureaucracy in Indonesia. In Chapter 6, governmentality theory on neo-liberalism and welfare liberalism will be combined with theories of neo-liberal conservation (or economic rationalism) to analyse REDD+ incentives and carbon markets. Chapter 7 will apply governmentality theory on advanced liberalism to theories of deliberative democracy, providing a framework to evaluate participation and engagement programs. Chapter 8 utilises Foucault s theories of power and knowledge, in addition to bio-politics, discipline, sovereignty and security to conduct a more in-depth analysis of the role of state and non-state agents. 101

122 Admintration and regulations Liberalism/biopolitics Bureauracratic failure Aministrative rationalism Control, accountability Experts and responsible communities Technical science Government performance, sustainable management Government policy or program Governmentality Problematisation Rationality Technologies Subjectivities Form of Knowledge Discourses Markets and incentives Neo-liberalism/welfare Market failure Economic rationalism Self, performance Price motivated individual Cost/benefit analysis Win-win development and conservation Participation and engagement Advanced liberalism Legitimacy failure Deliberative rationalism Steering, agency Democratic citizen Multiple epistemologies Stakeholder rights and program effectiveness Figure 4.2 Combined Theoretical Framework Building this theoretical framework has several benefits. First, the approach integrates broader macro-processes of rule with the micro practices of governing, enabling a focus on the localised circumstances and connections between rationalities and technologies, and the visibilities and subjectivities formed (Bröckling et al. 2010, p.23). It provides an organisational framework and a clear set of concepts to compare and contrast the national and sub-national programs, the interrelationships between the reform components, and importantly how different actors engage in the design and implementation of REDD+. Second, the current categories of governmental technologies found in the literature are limited in their application to contemporary political and social phenomena. An aim of this research, therefore, is to expand the theoretical scope of technologies and the relationship with specific policy instruments and forms of knowledge. The research will explore how different technologies function in terms of governing people and populations, their methods of operation, and how they change, steer and inform various activities. The aim is to understand how technologies create and construct certain subjectivities around the people who govern, as well as those who are governed. 102