SUBNATIONAL GOVERNANCE AND CONFLICT The Merits of Subnational Governance as a Catalyst for Peace Stefan Wolff, Simona Ross, and Asbjorn Wee Subnational Governance and Conflict The Merits of Subnational Governance as a Catalyst for Peace Stefan Wolff, Simona Ross, and Asbjorn Wee © 2020 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundar- ies, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemina- tion of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202- 522-2625; e-mail: pubrights@worldbank.org. Cover photo: © Foad Al Harazi/World Bank. Contents List of Case Studies vii List of Abbreviations ix Acknowledgments xi Preface xiii Background Studies xv Executive Summary xvii 1. Introduction 1 Concepts and Definitions 2 Intrastate Conflict with Subnational Dimensions 2 Political Settlements 3 Subnational Governance Arrangements 4 Institutions as Packages 4 Study Approach and Organization 5 Notes 7 2. Subnational Governance and Conflict 9 Drivers of Intrastate Conflict with Subnational Dimensions 11 Onset 13 Duration 17 Recurrence 17 From Drivers to Mitigation 18 Subnational Governance 19 Governance Packages 19 Forms of Subnational Governance 20 Subnational Governance to Mitigate Conflict 23 Conflict Drivers Mitigated through Subnational Governance 24 Political Exclusion 24 Economic Exclusion 24 Social Exclusion 24 Cultural Exclusion 24 iii Security 25 Notes 27 3. Framing Subnational Governance to Secure Peace 29 Subnational Governance to Share Political Power 29 Power Sharing and Subnational Politics 32 Subnational Governance to Redistribute State Resources 36 Sharing the Benefits of Natural Resources 38 Ensuring the Provision of Adequate Financial Resources 39 Subnational Governance to Extend State Authority 44 Subnational Governance to Strengthen State-Society Relations 49 Subnational Governance as a Slippery Slope to Secession? 56 What’s in the Law—Political Settlement Provisions 63 Legal Provisions That Reflect Power Balances and Local Realities 64 Constitutional Guarantees for Minorities and Marginalized Groups 65 Balancing Agreements between Ambiguity and Precision 65 From Paper to Peace 67 Addressing Commitment Problems 71 What Role for Third Parties? 73 Implementation Is Key 73 4. Strengthening Subnational Governance in Political Settlements 79 Three Conclusions: Process, Content, and Context 80 Process 82 Inclusiveness 82 Conclusiveness 82 Content 83 Match of Institutional Arrangements to Conflict Drivers 83 Appropriate Dispute-Resolution Mechanisms 84 Constitutional Entrenchment and Other Guarantees 85 Sufficient and Sustainable Financing 85 Context 86 Parties’ Acknowledgment of Each Other’s Right to Be Part of the Common State with Their Distinct Identities 86 Parties’ Efforts toward Implementation 86 iv Contents Absence of Effective Domestic Spoilers 87 Absence of Effective External Spoilers 87 Matching Institutional Bargaining to the Effective Balance of Power 88 International Support for Design, Implementation, and Operation 89 5. Conclusion 91 Appendix I. Case Studies Matrix 93 References 97 About the Authors 107 Contents v List of Case Studies Case Study 1  Kenya—Reducing the Stakes of Winner-Takes-All Politics 33 Case Study 2 The Philippines—Exploring the Merits of Accommodation through Asymmetrical Fiscal Autonomy 40 Case Study 3 Somalia—State Collapse and Loss of Monopoly on the Use of Force 46 Case Study 4  Myanmar—Renewing the Social Contract 53 Case Study 5  Bosnia and Herzegovina—Administrative Walls of Distrust 60 Case Study 6  Yemen—A Flawed Constitutional Framework 67 Case Study 7  Mali—A History of Unmet Expectations 74 vii List of Abbreviations AMISOM African Union Mission in Somalia ARMM Autonomous Region in Muslim Mindanao CAB Comprehensive Agreement on the Bangsamoro CDC Constitutional Drafting Committee FCV fragility, conflict, and violence GDP gross domestic product KANU Kenya African National Union MILF Moro Islamic Liberation Front MNLA National Movement for the Liberation of Azawad MNLF Moro National Liberation Front NCA Nationwide Ceasefire Agreement NDC National Dialogue Conference NLD National League for Democracy ODM Orange Democratic Movement PNU Party of National Unity SDG Sustainable Development Goal UN United Nations UNDP United Nations Development Programme USAID U.S. Agency for International Development ix Acknowledgments This study would not have been possible without the valuable support of our col- leagues at the World Bank’s Fragility, Conflict, and Violence (FCV) Group. This work was prepared under the direction and leadership of Alexandre Marc, Chief Specialist for FCV. The study’s findings are drawing from seven country case studies, and we are particularly grateful to our case study authors: Christine Wallich for Bosnia and Her- zegovina; Karuti Kanyinga and Prisca Kamungi for Kenya; Nikolas Myint and Corey Pattison for Myanmar; George Anderson for Yemen; and Samuel Chittick and Kai Kai- ser for the Philippines. The case study on Somalia was prepared by Simona Ross, and the case study on Mali was prepared by Asbjorn Wee, while the study on Bosnia and Herzegovina benefited from major contributions from Argyro Kartsonaki. We would also like to express our gratitude to each country’s Country Management Unit for its comprehensive review of the case studies and for providing critical insights. Special thanks to our peer reviewers Deborah Isser, Guenter Heidenhof, Balakrishna Menon Parameswaran, and Ingo Wiederhofer. We are also grateful for the inspiration and guidance we received from Yongmei Zhou, Edouard Al-Dahdah, and Michael Woolcock; valuable feedback also came from Douglas M. Addison. We also thank Lauri Scherer for her excellent editorial support. We are also, of course, very grateful to Shep- herd Incorporated for their editorial work and valuable support with the design and finalization of the publication. xi Preface In recent years, the World Bank has made it a priority to address the drivers of fragility, conflict, and violence as a prerequisite to achieving its goal of ending poverty and boosting shared prosperity. Today, a substantial share of the world’s population lives in contexts where sustainable development is stymied by the effects of violent conflict. With a shift in the nature of conflicts that are fought within state borders but involve a multiplicity of actors, building consensus on subnational governance arrangements that shape the distribution of political power and state resources between the center and the periphery is key to preventing, mitigating, and resolving conflict. The potential for subnational governance to serve as a catalyst of peace has been widely recognized. It is reflected in the frequent inclusion of subnational governance arrangements in political settlements. Following the events of the 2011 Arab Spring, many countries that are transitioning from war to peace and from authoritarianism to democracy are exploring different state models that entail some form of subnational governance. The study seeks to analyze the merits and risks of subnational governance as a con- flict mitigation mechanism within the framework of political settlement processes. This work is a contribution to the broader effort to close the knowledge gap on how development interventions can support the transition from war to peace. This study builds on the findings of the World Development Report 2011: Conflict, Security, and Development and the World Development Report 2017: Governance and the Law. It also provided critical input for and was informed by the joint United Nations–World Bank study, Pathways for Peace: Inclusive Approaches to Preventing Violent Conflicts. The findings of this study are largely informed by in-depth analysis of seven case studies of countries where subnational governance arrangements have been applied and proposed as a tool for mitigating violent conflict. These include Kenya, the Philip- pines, Somalia, Myanmar, Bosnia and Herzegovina, Yemen, and Mali. While there is broad recognition regarding the shortcomings and potential risks of using subnational governance reform as a mechanism for conflict mitigation, it is also clear that there are few, if any, credible alternatives for governing plural societies and promoting peaceful dispute resolution in fragile and conflict-affected contexts. xiii Background Studies Anderson, George. 2016. Yemen: A Flawed Constitutional Design. Case study for World Bank Study, Subnational Governance and Conflict. Washington, DC: World Bank. Chittick, Samuel, and Kai Kaiser. 2016. The Philippines: Exploring the Merits of Accommodation Through Asymmetrical Fiscal Autonomy. Case study for World Bank Study, Subnational Gov- ernance and Conflict. Washington, DC: World Bank. Kanyinga, Karuti, and Prisca Kamungi. 2016. Kenya: Reducing the Stakes of Winner-Takes-All Politics. Case study for World Bank Study, Subnational Governance and Conflict. Washington, DC: World Bank. Myint, Nikolas, and Corey Pattison. 2016. Myanmar: Renewing the Social Contract. Case study for World Bank Study, Subnational Governance and Conflict. Washington, DC: World Bank. Ross, Simona. 2016. Somalia: State Collapse and Loss of Monopoly on the Use of Force. Case study for World Bank Study, Subnational Governance and Conflict. Washington, DC: World Bank. Wallich, Christine. 2016. Bosnia and Herzegovina: Administrative Walls of Distrust. Case study for World Bank Study, Subnational Governance and Conflict. Washington, DC: World Bank. Wee, Asbjorn. 2016. Mali: A History of Unmet Expectations. Case study for World Bank Study, Subnational Governance and Conflict. Washington, DC: World Bank. xv Executive Summary Study Background and Rationale The world is experiencing a resurgence in violent conflicts. Most of these conflicts are fought within state borders, often within the context of political transitions, where con- flict parties seek to renegotiate and determine the nature of their state and the corre- sponding intergovernmental system. Preventing and resolving such intrastate conflicts requires a better understanding of the merits and risks associated with subnational gover- nance arrangements, which fundamentally shape how political power and state resources are shared and divided between the center and the periphery of countries. The past few years were marked by one of the most violent eras in recent history. The new century began with an ebb in global conflicts, compared to the 1990s. Yet, since 2010 there has been a rapid increase in violent conflicts, with more countries affected at any time since 1993. Such conflicts have resulted in a wave of forcibly dis- placed people, high numbers of casualties, and a record level of incidents of terrorist attacks. Since the end of the Second World War, global conflict dynamics have also experi- enced a general shift away from interstate conflicts and toward intrastate conflicts. With this shift come many changes regarding the nature of conflict, the drivers of instability, and the possible mechanisms with which to mitigate the risk of violence. Some intrastate conflicts have subnational dimensions and feature tensions between different identity groups. These types of conflict have been shown to be frequent, vio- lent, and long lasting. Moreover, intrastate conflicts with subnational dimensions are likely to spread beyond the state in which they start and draw in neighbors and regional powers, as evidenced by the conflicts in Bosnia and Herzegovina, Mali, Somalia, and Yemen. Intrastate conflicts with subnational dimensions are among the most pro- tracted conflicts, and consequently they have significant and damaging effects on states’ abilities to meet the needs of their citizens. Intrastate conflicts with subnational dimensions frequently manifest themselves in demands for greater levels of subnational self-governance that clash with efforts by the cen- ter to retain control over policy making. Subnational governance arrangements—which range from administrative decentralization to federalism and are used on their own and in xvii combination with other appropriate governance mechanisms—have the potential to miti- gate intrastate conflict with subnational dimensions by offering a compromise between competing demands for greater subnational governance and centralization. Despite a wealth of literature on this topic, there remains a lack of knowledge and guiding principles regarding how development actors can facilitate the design, imple- mentation, and functioning of subnational governance arrangements within political settlement processes in fragile and conflict-affected states. To achieve effective engage- ment, development actors need a better understanding of which conflict drivers respond to subnational governance arrangements and which complementary mechanisms make mitigation efforts sustainable. While existing literature emphasizes the importance of taking a process-oriented approach to political settlements, it does not offer an inte- grated approach for how development actors can support such processes. This knowledge gap creates a unique opportunity for this study to contribute to more effective engagement of international development actors with subnational gov- ernance arrangements, not only in the context of conflict settlement but also in light of findings in the joint United Nations–World Bank study Pathways for Peace highlight- ing that decentralization is also a critical component of successful conflict prevention. Approach To reduce this knowledge gap and provide development actors with operational guid- ance, this study explores the potential of subnational governance arrangements to mit- igate conflict. The analysis is based on a systematic review of existing literature on the drivers of and mitigation measures for intrastate conflicts. It also draws on lessons from seven country case studies that examined development actors’ support for polit- ical settlements with subnational governance dimensions. The seven case studies explore conflict dynamics and mitigation processes in Kenya, the Philippines, Somalia, Myanmar, Bosnia and Herzegovina, Yemen, and Mali. The case studies provided the empirical basis for a structured, focused comparison of experiences with strengthen- ing subnational governance in fragile and conflict-affected settings. Based on this approach, the study derived a series of recommendations for development interven- tions in political settlement processes in which subnational governance arrangements have been identified as key conflict-mitigating mechanisms. Organization This study is structured across five sections. Following an introduction that defines key concepts and discusses the study’s approach, Chapter 2 offers a systematic review of the existing literature on subnational governance and conflict and explores the merits of subnational governance arrangements in addressing relevant conflict drivers. Chap- ter 3 elaborates on a set of emerging themes that require closer examination to analyze the linkages between subnational governance and conflict. Chapter 4 applies these insights to the analysis of individual country case studies, summarizing and compar- ing relevant findings. Chapter 5 draws the study’s main conclusions. xviii Executive Summary Key Findings Among the critical takeaways from this study are that political settlement processes are always highly politicized; this is because they reflect and institutionally shape the power balances between different political actors. Development partners involved in such processes must understand that their engagement—however neutral it may have seemed in its conception—has an impact on how the domestic political context is reshaped. At its core, subnational governance is concerned with the distribution of political power. Thus, applying subnational governance arrangements as conflict miti- gation or prevention mechanisms is fundamentally about redesigning political power-sharing and power-dividing arrangements and redistributing state resources. Generally, political power and state resources are consolidated at the center and dis- persed according to subnational governance arrangements to the periphery. This is based on the assumption that the central government has the legitimacy and capacity to exert state authority in the periphery. Broad acceptance of state legitimacy and authority rests on the pillars of strong state-society relations, which presumes a state’s capacity to provide basic services at the local societal level. As a conflict mitigation mechanism, agreed-upon subnational governance arrange- ments are often entrenched in a political settlement’s legal framework. However, given the power shifts that result from redistributing political power and state resources, political settlements that feature subnational governance arrangements must address potential commitment problems among the conflict parties. When governance arrangements seek to shift power away from the center to lower levels of governance, there will be winners and losers among the political elite. Consequently, each process of formulating a new political settlement that involves subnational governance arrangements will engender various degrees of both opposition and support. A sustainable peace is one that is owned locally and includes, to every extent possi- ble, all conflict parties, while also reflecting the actual power balances between them. Conflict prevention measures and political settlements require active implementation; they hardly ever “self-execute.” International development actors need to play their part to avoid implementation delays and failures, including by providing security guarantees and helping—with both technical expertise and financial resources—to build local capacity and ensure the functioning of institutions as agreed on in a settlement. The extent to which subnational governance arrangements can successfully con- tribute to sustainable peace, democracy, and development depends on a number of factors. While these factors are determined by local contexts, this study identifies a set of lessons from the case studies along three dimensions: (1) the process of designing, implementing, and ensuring the functioning of subnational governance arrangements; (2) the content of the institutional package within which subnational governance arrangements are embedded; and (3) the domestic and external context of fragile and conflict-affected states. Regarding the process, the study finds that, in line with the political settlements perspective, inclusive and conclusive negotiation processes are more likely to produce sustainable outcomes for peace and development. Unless all powerful actors are included and have the opportunity to shape future institutional arrangements, any Executive Summary xix negotiated outcome is unlikely to adequately reflect the effective balance of power. A lack of inclusiveness, consequently, will result in an unsustainable political settlement that is incapable of achieving the required political stability, because it will “produce” effective spoilers. Moreover, there is a need to address all critical drivers of conflict in the negotiation process. This does not imply that all the critical drivers be resolved as part of negotiations, but rather that they must be recognized as issues that need to be addressed as part of the broader process of bargaining and formulating a new political settlement. Some contentious matters may need to be postponed, while others may need to be delegated to specific bodies, either to work out a technical (rather than political) solution or to arbitrate in a particular dispute. There are four crucial elements concerning the content of any agreement regarding the design of the institutions, processes, and mechanisms that form the institutional package for resolving differences between parties within a new political settlement. These include: • Whether institutional arrangements match conflict drivers. Since not all drivers are equally responsive to conflict mitigation through the application of subnational governance arrangements, it is important to look beyond what conflict parties advocate and to establish whether what they promote addresses particular conflict possible— drivers. In addition, it is crucial that a settlement be as comprehensive as ­ in other words, that it deals with all identified drivers of conflict, which in turn requires that institutions fit with each other as a “package” and consider possible tensions between already established customary governance systems and new, for- mal arrangements that are introduced as part of the settlement. • Appropriate dispute-resolution mechanisms. Engineered political settlements nor- mally feature a degree of (constructive) ambiguity in their provisions; this allows negotiators to claim that they achieved what they promised their constituents, and in turn helps keep them on board during the implementation process. How- ever, doing so often creates situations in which disputes arise over parties’ inter- pretation of particular provisions during a settlement’s implementation phase. Moreover, disputes may also arise later on—for example, between authorities at different levels of government, such as a subnational and a central admin- istration. To mitigate the potential for such disputes to derail a peace process or destroy a political settlement, it is important that parties agree to dispute- resolution mechanisms as part of the political settlement. These mechanisms will subsequently enable them to resolve differences by political means rather than by renewed armed conflict. • Constitutional entrenchment and other guarantees. Constitutional entrenchment is critical for a political settlement’s sustainability. Constitutional and other guaran- tees send a clear signal to all parties and their constituents about the commitment they are making to implement and uphold the settlement, while leaving open the possibility of future revisions by broad consensus. Building guarantees into formal legal structures and informal institutions also create incentives for parties to con- structively engage in negotiations, because guarantees make it more likely that any institutional arrangements agreed upon will in fact be implemented and operated. xx Executive Summary • Sufficient and sustainable financing. Political settlements remain only commitments on paper until they are implemented. While implementation is first and foremost a question of the parties’ political will, it also depends on the availability of resources. Especially in cases involving subnational governance arrangements, it is critical to ensure that resource institutions are created (where they do not already exist) and sustainably financed, so they can function effectively and deliver public goods and services. It is also critical to establish financing mechanisms such that old inequalities are not exacerbated and new ones are not introduced. As such, revenue availability must match the subnational governance arrangements, which is particularly impor- tant for countries that have for years (or even decades) been held captive by violent conflict and subsequently lack fiscal resources and institutional capacity. With respect to the context, the institutions, processes, and mechanisms that consti- tute any political settlement do not operate in a vacuum. Rather, they exist in a politi- cized and volatile environment that is characterized by high levels of distrust among the presumptive parties and very real risks posed by spoilers that could derail a settlement process. While settlement design—namely the content dimension of success—can address some of these issues, both the domestic and external contexts in which a settle- ment is designed, implemented, and functions require active “management,” including by international development partners. • Domestically, international development actors need to focus on enabling and sus- taining a minimum degree of consensus between parties regarding their respective right to be part of a common state with their distinct identities. This should be undertaken through joint efforts by parties at the central level, for instance, to man- age a constitutional reform process, and it can also apply to individual efforts, such as delivering governance at a subnational level in line with a political settlement and/or a constitutional order. Doing so minimizes the potential for domestic spoil- ers to emerge and derail a new political settlement. • Externally, international development actors should focus on limiting the likeli- hood that external spoilers will emerge or be effective in derailing the peace and/or reform process. Development actors should also acknowledge that while local own- ership of a political settlement is critical, it is not by itself sufficient; however, it can be enhanced by the right kinds of international support. Executive Summary xxi CHAPTER 1 Introduction The world is experiencing a resurgence in violent conflict. Many of these conflicts are fought within state borders, often in the context of deeply divided plural societies that are undergoing political transitions to determine the nature of the state and how sub- national governance arrangements should redistribute political power and state resources. Resolving intrastate conflicts requires a better understanding of the merits and risks of subnational governance arrangements, which fundamentally shape the redistribution of political power and state resources between the center and the periphery. The past few years were marked by the highest level of violent conflict in the post– Cold War era. Countries in Sub-Saharan Africa, the Middle East, and parts of Asia were disproportionally affected by the rise in violence. Above all, the consequences of these conflicts are mostly shouldered by civilian populations, a fact that is tragically illustrated by several of the case studies. This is evident in Yemen, for example, where the ongoing conflict has caused an estimated 24.1 million people, out of a total popu- lation of 30.5 million, to suffer from famine and need humanitarian assistance (OCHA 2019). Two of the key drivers of this conflict in Yemen are a historic struggle over polit- ical exclusion and a failure to negotiate viable subnational governance arrangements. Over the course of the past century, global conflict dynamics experienced a sub- stantial shift away from predominantly interstate conflicts toward intrastate conflicts. Since the end of the Second World War, conflicts between states have almost disap- peared, while conflicts fought within state borders have proliferated. Following a surge of civil wars in the early 1990s, the number of intrastate conflicts decreased at the beginning of the new century, only to rise again in recent years (Pettersson and Wal- lensteen 2015). Contributing to the intractability of intrastate conflicts is their multi- dimensional nature and the fact that numerous state and non-state actors are involved (United Nations and World Bank 2018), which means that many intrastate conflicts are in fact internationalized.1 This shift in conflict dynamics stresses the need to rethink governance structures within state borders in ways that both reflect the power balance among various politi- cal actors and strengthen relations between the state and society, particularly with groups that have traditionally been marginalized. Internal governance structures must 1 ensure access to representation and participation in political decision making; this helps bring governance closer to the people and provides mechanisms to address grievances and to peacefully resolve disputes. Subnational governance arrangements are critical for determining relations between different levels of government and to define citizens’ role in participating in government through civic exercises at these levels. Subnational governance arrangements have become a central element of political set- tlements across a wide range of conflicts. In intrastate conflicts, where groups strive for greater self-governance or for participation in policy making at the center, subnational governance arrangements are often default mechanisms of settlement processes. Yet, even governance are not the main factor, some degree of decen- in cases where claims for self-­ tralized power from the center to lower levels of governance often makes up part of a broader political settlement. Such arrangements aspire to improve the quality of gover- nance by broadening opportunities for citizens’ representation and participation in a democratic framework of institutions. Questions pertaining to the structure of the state and the redistribution of political power and state resources across multiple layers of gov- ernance are at the heart of institutional design and constitutional reform processes in many of the world’s fragile and conflict-affected states. Therefore, it is essential to get sub- national governance arrangements right in these contexts; doing so is critical for these countries’ prospects for long-term peace, democracy, and sustainable development. Realistically however, subnational governance arrangements are not a panacea for peace building, particularly not on their own. There is no one-size-fits-all form of sub- national governance arrangement that can foster stability in every context. The question may no longer be whether subnational governance arrangements are a useful mechanism of conflict mitigation, but rather which approach domestic stake- holders and their international development partners should take. Subnational gover- nance arrangements, both in a substantive sense—that is, which subnational and intergovernmental institutions should be established, and in a procedural sense—how, by whom, and in which order these institutions should be designed and implemented, need to be examined. It is this debate that the following study addresses. Concepts and Definitions This section sets out a terminologically consistent and intellectually coherent framework for the key concepts and definitions used throughout this study, namely intrastate conflict with subnational dimensions, political settlements, and subnational governance arrange- ments. In this context, it also elaborates on the notion of institutions as packages. Intrastate Conflict with Subnational Dimensions Our definition of intrastate conflict with subnational dimensions draws on the defini- tions put forth by the Uppsala Conflict Data Program. Intrastate conflict is understood as an “armed conflict . . . where the use of armed force between two parties, of which at least one is the government of a state, results in at least 25 battle-related deaths in one calendar year.”2 Subnational dimensions are taken to have a similar meaning in the Uppsala Conflict Data Program’s use of “territorial incompatibility”—namely, a 2 Subnational Governance and Conflict contested incompatibility “concerning the status of a specified territory” (Uppsala Conflict Data Program 2015). In a number of cases, there is overlap with the category of “government incompatibility,” as an “incompatibility concerning the type of politi- cal system, the replacement of the central government or the change of its composi- tion.” It is also conceivable that intrastate conflicts do not start out featuring territorial incompatibility, but that such tensions develop over time and that their settlement includes subnational governance arrangements. Political Settlements Academic and policy literature have widely used the concept of political settlements to describe the broader political context in which development is pursued, including with the support of the international donor community.3 In the context of war-to-peace transitions, the concept has been used to emphasize the political foundations of sus- tainable peace. As a result of its increasing use across disciplines and contexts, the con- ceptualization and operationalization of political settlement varies. A political settlement’s perspective is central to this study since it captures the complex and context- dependent nature of contemporary approaches to addressing intrastate conflict. For the purposes of this study, we adopt Khan’s (2010, 21) definition of political settlements, which considers them to be “an interdependent combination of a struc- ture of power and institutions at the level of a society that is mutually ‘compatible’ and also ‘sustainable’ in terms of economic and political viability.”4 The latter two condi- tions require sufficient economic performance to avoid economic crises, and require levels of conflict and violence be below a threshold at which core institutions begin to unravel (Khan 2010). Implied in this definition is Khan’s (2010) contention that polit- ical stability and economic performance are functions of interaction between the structure of institutions and the distribution of power between key actors, including between elites and their constituents. Put differently, “a political settlement describes a sustainable and compatible combination of power and institutions” (Khan 2010, 25) that generates and distributes benefits in such a way that no actor emerges powerful enough to challenge the existing formal or informal institutional structure. Political settlements often introduce new governance systems that reshape the distribution of political power and state resources between actors in the center and the periphery. This perspective on the interaction between power and institutions is particularly critical at key moments in a country’s history when political settlements are redefined, such as during war-to-peace transitions when a new political settlement emerges through a process of negotiation between relevant actors in a context shaped by domestic and external actors. This is reflected in the assertion of Bell (2015, 5) that political settlements are best conceived of as “a project of engagement with the possi- bilities of transformation in situations where powerful and violent actors appear to hold all the cards, rather than a project of designing a ‘thing’ called a political settle- ment.” In this sense, a political settlement is a dynamic process of negotiating, imple- menting, and operationalizing a peace agreement, rather than just a reference to the agreement itself. This study’s focus is on a specific type of political settlement, which Whaites (2008, 15) calls engineered settlements, described as those that are “explicitly Introduction 3 negotiated sometimes with external mediation, often as part of a formal ‘peace process’ or following external intervention.” Political settlements bridge the two concepts of state building and peace building,5 and encourage a long-term perspective regarding peace processes. Political settlements entail negotiations that define “the field on which politics is played, [by] setting out the nature and rules of the ‘game’ and [by] setting conditions that determine which players are ‘in’ and playing what role” (Barnes 2009, 9). These are followed by implementing and operating the agreements reached. From this perspective, political settlements are packages of institutions whose long-term success depends on their negotiation, imple- mentation, and functioning (process dimension), on the institutional structures pre- scribed (content dimension), and on the extent to which they reflect the real distribution of power among critical interest groups (context dimension). A p ­ articularly important aspect of “designing” ­ institutions—and the processes of their negotiation, implemen- tation, and ­ functioning—is the degree to which they include all relevant groups that are part of the political settlement in question (Castillejo 2014; Hickey 2013; Rao 2013; Rocha Menocal 2015). Subnational Governance Arrangements For the purposes of this study, the subnational governance arrangements of interest involve the institutions set up as part of a political settlement that address questions of territorial and intergovernmental state structures and conflict settlement. The defini- tion adopted follows Wolff (2013a) and delineates subnational governance arrange- ments as the legally entrenched power of territorially delimited entities within the internationally recognized boundaries of existing states to exercise public policy func- tions independently of other sources of authority in this state but subject to its overall legal order.6 This broad conceptualization includes five empirical types of subnational governance arrangements: confederation, federation, federacy (or autonomy), devolu- tion, and local self-governance. Closely related to subnational governance arrangements are the concepts of decen- tralization, delegation, and deconcentration, which pertain to the implementation of the principle of subsidiarity. Decentralization is often seen as an umbrella term that covers delegation (a more extensive form of decentralization) and deconcentration (a weaker form of decentralization); it refers to both the process and result of structuring a system of governance. In such arrangements, multiple layers of governance share authority and have the capacity to deliver public goods and services. Decentralization, delegation, and deconcentration seek to achieve effective and responsive governance in general, rather than to mitigate conflict specifically. That said, their basic principles can be applied to subnational governance arrangements used for conflict mitigation but may require certain adjustments. Institutions as Packages Institutions do not operate in isolation from each other; they must be viewed as “pack- ages” in which individual components interact, often in complex ways.7 When consid- ering the merits of subnational governance arrangements for conflict mitigation, we 4 Subnational Governance and Conflict need to analyze the process by which the arrangements are negotiated, implemented, and operated in the context of the broader institutional structures. There are a number of institutions that have been identified as particularly significant for the sustainability of subnational governance arrangements, including (1) fiscal decentralization, which awards subnational spending powers and possibly independent revenue-raising com- petences; (2) the structure at the center of the political system, which determines opportunities for political representation and participation of groups that benefit from subnational governance arrangements;8 (3) the structure of the political party system and the presence and strength of specific regional parties; and (4) the “spatial” arrangements for subnational governance, such as the number, demographic makeup, and economic strength of subnational units. Study Approach and Organization With the study’s conceptual framework established, we now elaborate on the overall approach that methodologically underpins it. A review of existing literature helps establish a link between subnational governance and conflict, highlighting potential merits and risks of subnational governance arrangements as conflict mitigation mech- anisms. The study identifies a number of reasonable assumptions about the utility of subnational governance as a way to address certain conflict drivers and the conditions under which peace can be sustained. The individual and comparative analysis of the case studies aims to illuminate two questions: (1) How can local, national, regional, and global actors make useful contri- butions to the negotiation and implementation of subnational governance reforms, and under what conditions? and (2) In light of past experience, what might be done differently to make their contributions more effective? The study seeks to review les- sons from when external actors have engaged in conflict mitigation efforts in the past and what impact their interventions have had on the negotiation, implementation, and operationalization of subnational governance arrangements. The analysis in this study is based on seven case studies—Kenya, the Philippines, Somalia, Myanmar, Bosnia and Herzegovina, Yemen, and Mali—that offer an in-depth exploration of each conflict and its settlement process. The cases selected were “histor- ical” and “contemporary” in nature, while covering a range of different c­ ontexts—from the more “traditional” war-to-peace transitions (Bosnia and Herzegovina, the Philip- pines) and postauthoritarian transitions (Myanmar, Yemen), to broader institutional and constitutional reform projects in the context of persistent or recent conflicts (Kenya, Mali), to countries facing multiple challenges related to subnational gover- nance (Somalia). Analyzing ongoing conflicts enabled us to conduct a form of action research that both allowed participant observation as an additional research method and facilitated more immediate practitioner feedback. The seven case studies do not represent an empirically representative measure but provide lessons and critical insights for comparative analysis based on a set of questions posed in each case study.9 In the Bosnia and Herzegovina case study, Christine Wallich examines the 1995 Dayton Peace Accords and the state inefficiencies they created as a trade-off for avert- ing secession and achieving an end to the war that had for three years devastated the Introduction 5 former Socialist Federal Republic of Yugoslavia. The case study of Kenya, explored by Karuti Kanyinga and Prisca Kamungi, is very different. In Kenya, control over territory was not at the heart of the 2007–08 postelection violence, but subnational governance arrangements were part and parcel of a constitutional reform process set in motion to settle (and prevent the resurgence of) interethnic conflict fought over political power and access to state resources. While the jury is still out on whether the constitutional reforms have effectively achieved this, Kanyinga and Kamungi observe two important trends: “There are signs that devolution of power and resources reduces the sense of marginalization with regard to national level intercommunity relations. [And] at the level of county governments, devolution arouses intra-group conflicts on account of how groups feel about distribution of development in the county.” This highlights the need for subnational governance arrangements to be designed so they are sensitive to their contexts. Interestingly, it also echoes a point that Wallich makes about the political impact of aid in Bosnia and Herzegovina, and one that Sam- uel Chittick and Kai Kaiser, authors of the case study on the Philippines, make about the importance of considering both national and subnational political economies of war and peace: that decentralization—namely, the creation of lower-level governance units—affects how resources are distributed and whether distribution is perceived as fair among communities in the same unit and between units, which in turn can spark new grievances or reignite old ones as much as it can address them effectively. Asbjorn Wee considers the case of Mali as an example of how a perennial failure to implement settlements with subnational governance arrangements has become a driver of recurring instability. In the case study on Myanmar, Nikolas Myint and Corey Patti- son analyze the political transition initiated by the enforcement of the 2008 Constitu- tion, which, after six decades of war, offers the first step from an authoritarian military regime toward a democratic governance system. The central government envisioned subnational governance as a strategy for state penetration in the periphery and to strengthen the social contract through improved local service delivery. Concerning the case of the Philippines, Chittick and Kaiser point out that the “key grievance is charac- terized as one of identity, with the proposed solution being a recognition of their funda- mental right to self-determination, addressed through agreement on a regional government with genuine autonomy.” Asymmetric fiscal autonomy was proposed to accommodate the minority groups inhabiting the region of Muslim Mindanao and to dampen secessionist aspirations. Simona Ross’s study on Somalia offers important insights into a situation in which state building in the context of a nearly complete absence of credible institutions and lack of consensus over the structure of the state itself represents “a major source of instability.” Finally, George Anderson examines the case of Yemen, where the ongoing conflict stems from a failure to design a subnational governance structure that grants its communities genuine political participation. The remainder of this study is presented in four parts. Chapter 2 elaborates on recent trends in the nature of intrastate conflicts, including more specifically conflicts with subnational dimensions, and the spectrum of subnational governance arrange- ments. Further, Chapter 2 explores the connection between conflict and subnational governance and the merits and risks of subnational governance arrangements as con- flict mitigation mechanisms. Chapter 3 provides an in-depth analysis of emerging 6 Subnational Governance and Conflict themes that have significant implications for subnational governance reform in the framework of political settlements and how these arise in the country cases. Chapter 4 draws a number of conclusions about the merits of subnational governance arrange- ments as a tool for the prevention, mitigation, and resolution of intrastate conflict and reflects on their implications for international development actors. Chapter 5 con- cludes the study by recapping its key findings and proposing a way forward. Notes 1. The Uppsala Conflict Data Program (2015) considers a conflict internationalized if “the government side, the opposing side, or both sides receive troop support from other govern- ments that actively participate in the conflict.” This definition is similar to what Lake and Rothchild (1998) term “escalation,” which “occurs when a conflict in one country brings in new, foreign belligerents.” Importantly, conflicts can also spread by diffusion—i.e., “when ethnic violence in one state increases the probability of conflict in a second” (23). 2. Other data sets use the higher threshold of 1,000 annual battle deaths (Collier and Hoeffler 2004; Fearon and Laitin 2003; Wimmer and Min 2006), thus excluding about half of all intrastate armed conflicts compared to using the lower threshold of 25. 3. Various literature reviews attest to the widespread and variable use of political settlements. See, for example, Di John and Putzel (2009), Jones et al. (2012), and Laws (2012). 4. The first use of the term by Khan (1995) describes political settlements as a country’s “inherited balance of power” (71). 5. The relationship between peace building, state building, and political settlements is dis- cussed, among others, in Department for International Development (2010), Gutierrez (2011), and Schünemann and Lucey (2015). 6. This definition draws on earlier work by Weller and Wolff (2005). 7. See, for example, Belmont et al. (2002), Hartzell and Hoddie (2003), Schneckener (2002), and Wolff (2011c). 8. These political arrangements at the center predominantly take the form of power sharing in postconflict contexts, as a form of governance arrangement specifically applied for the pur- poses of conflict mitigation and with the aim of increasing the political inclusion of previ- ously marginalized groups by ensuring their effective representation and participation in decision-making bodies, such as cabinets, legislatures, executive agencies, and judicial organs. Power sharing often accompanies, and may in fact be required by, the introduction of subnational governance arrangements in conflict contexts (Wolff 2009), but it is also considered a form of power sharing itself—territorial power sharing (Hartzell and Hoddie 2007). 9. These included: Which conflict causes could have been/were addressed by subnational gov- ernance arrangements? Were these arrangements correctly identified, or did they miss the target? What other governance arrangements were part of the negotiated agreement? To what extent and why were these arrangements implemented and sustained or revised after implementation? Which were the most pivotal factors, and who were the most pivotal actors in achieving this outcome? Introduction 7 CHAPTER 2 Subnational Governance and Conflict The world is more conflict ridden today than at any moment in the past decade (United Nations and World Bank 2018). From Yemen to Myanmar, from Somalia to Mali, intra- state conflict affects the lives of millions of people and has been proven to have staggering humanitarian and economic costs. There are an unprecedented 71.8 million forcibly dis- placed people worldwide (UNHCR 2018). The number of internally displaced people fleeing violent conflict and persecution doubled from 2016 to 2017 alone.10 In 2017, global violence and intrastate conflict came at an overwhelming economic cost of US$14.8 trillion.11 Today’s intrastate conflicts have resulted in an estimated 5.2 percentage points in lost gross domestic product (GDP) growth per conflict year (Lomborg 2013). The rapid deterioration of global peace has been explained by the inability to resolve ongoing conflicts. So not only is there an increase in the onset of intrastate conflict, there have been few successful efforts to resolve intrastate conflicts in the past decade (Institute for Economics and Peace 2018). To mitigate the devastating conse- quences of violence and conflict, the international community is promoting efforts to resolve intrastate conflict through the negotiation, implementation, and operational- ization of new political settlements. To ensure that such settlements are designed to effectively and sustainably mitigate such violence, it is paramount to understand the drivers and underlying grievances that drive intrastate conflict. Many intrastate conflicts have subnational dimensions and are fought over how to distribute political power and state resources between the center and periphery, as well as among subnational actors. Ending intrastate conflict with subnational dimensions requires governance interventions that address the root causes of such conflicts. When part of political settlements, subnational governance has been hailed as an effective measure to prevent, mitigate, and resolve intrastate conflict with subnational dimen- sions. Therefore, building, reforming, and strengthening subnational governance arrangements have been among the most common governance interventions in fragile and conflict-affected countries in recent years. International organizations are increasingly aware of the linkages between subna- tional governance and intrastate conflict. Within the framework of the Sustainable Development Goals (SDGs), the international community committed to supporting efforts that strengthen subnational governance and to using the efforts as a tool to 9 foster peace and stability. Under SDG 16, which seeks to “promote peaceful and inclu- sive societies for sustainable development, provide access to justice for all and build effective, accountable and inclusive institutions at all levels,” two subtargets focus spe- cifically on the role subnational governance can play in achieving these goals.12 The World Bank has recognized subnational governance and conflict as central themes in its prominent World Development Report series. Most recently, the World Development Report 2017: Governance and the Law finds that power-sharing institu- tions “that balance, divide, and share power reduce the incentives to engage in violence by increasing the benefit of security” (World Bank 2017, 4.11), as do redistributive institutions “that allocate and redistribute resources and resource rents” (World Bank 2017, 4.12). This builds on earlier findings in the World Development Report 2011: Conflict, Security, and Development, which notes that “the argument in favour of greater decentralization of power (for example, through federalist structures that hand significant autonomy to local representatives) is that it can avert centre-periphery eth- nic conflict, or secession” (World Bank 2011, 166). Even prior to that assertion, the World Development Report 1999/2000: Entering the 21st Century observed that “when a country finds itself deeply divided, especially along geographic or ethnic lines, decentralized and local governance provides an institu- tional mechanism for bringing opposition groups into a formal, rule-bound bargain- ing process” (World Bank 1999, 107–8). The argument is that, based on empirical evidence, subnational governance arrangements have the potential to mitigate conflict by offering (1) a path to national unity; (2) a political solution to civil war; (3) an instrument for deflating secessionist aspirations by conceding enough power to fore- stall their escalation into violence; and (4) a mechanism to be responsive to grassroot demands and enlist local support for central policies (World Bank 1999). This has been echoed by other national and international development institutions. The U.S. Agency for International Development (USAID) makes a similar point about the “conflict-reducing potential of decentralization,” describing it as “especially appeal- ing in countries where ethnic, linguistic, religious, and cultural groups are concentrated in distinct territories or regions” because “decentralization in these settings can accom- modate diversity by giving subnational officials in the regions the power to offer differ- entiated programs that respect local preferences and cultural practices,” which “provides assurances to minority groups that their priority concerns will be considered.” Thus, “strengthening the level of government or administration where minority groups have influence . . . can discourage the formation of secessionist movements” (USAID 2009, 22). Similarly, the United Nations Development Programme (UNDP) finds that “the potential of local governance for peacebuilding, statebuilding and recovery often remains neglected” (UNDP 2016, 15), emphasizing that “more effective local governance systems that are responsive to people’s needs and aspirations, inclusive and resilient to crises should play a decisive role in delivering improved quality of life for all, reducing inequal- ities and cementing a resilient state-society relationship based on a shared understanding of respective roles and responsibilities” (UNDP 2016, 17). Leading international organizations recognize the potential merits of subnational gov- ernance as a conflict mitigation mechanism. Thus, such arrangements feature promi- nently in political settlements that have been negotiated, designed, and implemented in 10 Subnational Governance and Conflict several fragile and conflict-affected states. In the post–Second World War era, as countries emerged from violent conflicts in their quest for independence and following the collapse of the Soviet Union, many have reached political settlements to determine their nation’s governance structure and intergovernmental system. It is critical to get the negotiation, design, and functioning of subnational governance arrangements right; negotiated settle- ments have been the primary mechanism for ending civil wars since the end of the Cold War. Since then, subnational governance reform has featured prominently in postconflict political settlements, as former adversaries seek to reduce the likelihood of renewed con- flict by sharing power and resources, or as central governments seek to dampen social unrest and secessionist aspirations in exchange for greater inclusion or self-governance. For instance, in the 1990s, Africa experienced a surge of power-sharing agreements that were mostly based on consociational power-sharing models (Cheeseman 2011; Mogaka 2017) that also included subnational governance arrangements. In recent years, the international community has substantially increased its support to subnational governance reform in fragile and conflict-affected states. It is estimated that the World Bank allocates around $3 billion each year to support community- driven development projects, which are used as tools to strengthen local governance from the bottom up (UNDP 2016). International support for subnational governance reform resulted in the adoption of such arrangements in over 80 percent of developing countries from the 1980s to 2008 (Crawford and Hartmann 2008). Drivers of Intrastate Conflict with Subnational Dimensions Violent conflict is driven by the grievances held by some segments of society, the oppor- tunities to express discontent through violence, the feasibility of engaging in violence in the context of state fragility and institutional weakness, and the presence of catalysts that increase the risk for violent conflict (World Bank 2011). From 1990 to 2005 alone, 50 conflicts in over 36 countries had subnational dimen- sions, including, most prominently, the accommodation of minorities in plural societ- ies (Roeder and Rothchild 2005). In addition, secessionist conflicts accounted for one-fourth of all intrastate wars waged since the end of the Second World War (Wim- mer, Cedermann, and Min 2009). Secessionist conflicts often have a destabilizing effect on the entire region, as “the number of unresolved boundary issues across the world is a further cause of persistent interstate tensions” (United Nations and World Bank 2018, 11). Intrastate conflicts with subnational dimensions are not only the most prevalent form of conflict, they are also the most intractable. Intrastate conflicts are mostly fought between minority groups and the central government over control of territory and are the least likely to conclude with a political settlement, since governments are not expected to accommodate demands of minority or secessionist groups for indepen- dence or greater regional autonomy (Walter 2003). After all, the sovereign territory of a state is finite, and granting secessionist groups independence threatens its very exis- tence. Nonetheless, Fearon and Laitin (2008) also find that “separatist or autonomy- seeking rebel groups” gain “de facto or significant regional autonomy provisions in Subnational Governance and Conflict 11 40 percent of the cases,” indicating that institutional designs that emerge reflect a bal- ance of power on the ground that fosters a territorial “compromise” between defenders and challengers of an existing political settlement, such as an incumbent government and a rebel movement. A single country may experience multiple intrastate conflicts with subnational dimensions that take varying forms, with some non-state actors demanding greater decentralization within the territory of a sovereign state, while other non-state actors are advocating for independence. Consider Myanmar’s conflict dynamics, in which the military junta was at one point fighting multiple parallel conflicts with well over a dozen armed ethnic groups, all claiming some form of subnational governance or secession. Intrastate conflict can create an unstable and chaotic environment in which parallel localized conflicts can proliferate. In some countries, intrastate conflict creates a secu- rity vacuum that allows illicit smuggling and trafficking networks to flourish across porous borders, which in turn can drive communal disputes as non-state armed actors compete for access to and control over the financially lucrative networks. Conversely, in some cases, conflict that is driven by local issues risks being elevated (Jackson 2016) and can exacerbate tensions at the national level, leading to all-out civil war. Regarding intrastate conflicts with subnational dimensions, Buhaug and Rød’s (2006, 325) study of local determinants of civil wars in Africa found that such conflicts were more likely in areas “closer to the border, further away from the capital, petro- leum fields, and diamondiferous areas; they are less populated, less rugged, have less developed road networks, and are more likely to contain a minority language.” Simi- larly, Fearon (2004, 297–8) points out that so-called “sons-of-the-soil” conflicts typi- cally last longer than conflicts over government (coups or revolutions) because “a stable regional autonomy deal is harder to construct when the political centre’s stakes in the region are greater . . . [and] when the rebel force can extract more from a region during the course of a war.” This shift in conflict dynamics has been accompanied by a change in the character- istics of conflict drivers. The risk of violent conflict is strongly determined by how political power is distributed, which is reflected in the level of access that marginalized groups have to central state authorities (World Bank 2017). In authoritarian or repres- sive regimes, peripheral non-state actors typically initiate violence to contest the cen- tral government’s authority over their territory. Intrastate conflicts are also increasingly fought over state-society relations and the role that governance plays in determining these (Brinkerhoff 2011). This link has been most evident in the wave of instability that took place during and after the Arab Spring. Recognizing the centrality of institutions and governance systems in questions of conflict and peace, the World Development Report 2017: Governance and the Law con- cluded that violent conflict occurs when governance institutions break down and can no longer serve as a platform for the peaceful resolution of disputes. There is emerging consensus that strong institutional capacity is a key element of resilience against inter- nal stresses and external shocks that, in the absence of such capacity, have the potential to trigger violent conflict (World Bank 2011). 12 Subnational Governance and Conflict Intrastate conflicts with subnational dimensions take many forms, and their under- lying causes are multilayered and interconnected (Colletta and Oppenheim 2017). A set of common, overarching themes emerge from both the existing literature and our case studies regarding intrastate conflicts with subnational dimensions. This form of intrastate conflict occurs when parties contest: (1) the nature of the state and the gov- ernance structure that establishes the distribution of political power and state resources, particularly postconflict or during postauthoritarian transitions; (2) the subnational governance arrangements that determine this distribution between the center and the periphery; (3) the historical, political, economic, and social exclusion of marginalized groups; (4) the forceful integration of identity groups into sovereign state territory that is dominated by other groups; (5) demands for self-determination; and (6) communal dynamics and disputes over economic resources that lead to local- ized conflict. These intrastate conflict drivers are also likely to play out differently in three important ways: their onset (the conditions under which intrastate conflicts start), their duration (the factors that determine how long they last), and their recurrence (the circumstances under which resolved conflicts start anew).13 These are useful distinc- tions that help shed light on which causes must be addressed in a political settlement and which might respond to mitigation by subnational governance arrangements at a particular point in the conflict cycle. For example, issues that start conflict are often different from those that determine duration and affect recurrence. Onset These drivers operate at different levels of “causation” of intrastate conflict. Economic and identity-based marginalization are likely to cause grievances, motivating groups and their members to engage in conflict. Control over resources and rent seeking can either finance conflict or serve as a motivation in itself. A quest for self-determination, whether expressed as a demand for more self-governance or outright secession, is more likely to be of an instrumental purpose—such as to address an underlying prob- lem like economic or identity-based marginalization—than be an end in itself.14 Both motivation and feasibility determine what becomes a driver of intrastate con- flict. In other words, warring parties must be both willing and able to engage in con- flict. From a political settlements perspective, this means that the institutional structures no longer reflect the interests of groups in a given society that have the capa- bility to challenge the defenders of an existing political settlement, therefore also indi- cating a shifting balance of power between groups. Brown (2001) made an early attempt to differentiate between motivation and feasibility. He recognized a distinc- tion between underlying and proximate causes, the latter being the catalytic factors that transform potentially violent situations into actual deadly conflicts. There are four types of proximate causes: elite-triggered, mass-triggered, internally driven, and exter- nally driven. Internal mass-level factors—­ “bad domestic problems”—include rapid economic development or patterns of political and economic discrimination. In con- trast, internal elite-level factors—“bad leaders”—include a variety of power s ­ truggles, such as those that are encountered during transition processes. Among the externally Subnational Governance and Conflict 13 triggered factors, those at the mass level include spillover effects in so-called “bad neighborhoods,” such as the sudden, massive influx of refugees, while those at the elite level—“bad neighbors”—refer to deliberate actions undertaken by neighboring gov- ernments, such as supporting militant opposition groups (Brown 2001, 579–80). Proximate causes are part of a causal chain that link a conflict cause to its actual incidence. Analyzing proximate causes can yield useful insights into possible interven- tion points and strategies designed to disrupt the causal chain. While this may be an effective approach in the context of short-term crisis management, it is no substitute for more comprehensive and profound conflict mitigation that seeks to address the underlying causes of conflict. Social, political, and economic exclusion are the most commonly identified sources of grievances that, absent satisfactory redress, can serve as motivations for violent con- flict. Competition over resources that the state controls and distributes, including through the provision of public goods and services, can spark conflict because it even- tually creates a security dilemma for the groups that both control and challenge the incumbent regime. While the security dilemma explanation has incurred some criti- cism (Jenne 2012), it is the most commonly accepted causal mechanism in both social-psychological and rational choice theories of intrastate conflict that link griev- ances to conflict onset (Horowitz 1985; Kaufman 2001; Kaufmann 1996; Lake and Rothchild 1996; Posen 1993). Findings in the literature concerning the type of grievance more likely to lead to conflict have been relatively consistent over time and map well onto economic and identity-based exclusion drivers. Empirical evidence suggests that by 1953 only about 15 percent of conflicts were driven by ethnic diversity. This share had increased to nearly 60 percent by 2005. Subsequently, group mobilization along identity lines has been identified as the most important driver of civil war (Stewart 2010). The most coherent expression of this grievance-based argument is the notion of horizontal inequalities (Cederman et al. 2011; Stewart 2010). The core argument in this context is that “violent mobilisation is more likely when a group that shares a salient identity faces severe inequalities with regard to one or more of the following dimensions: economic, social, and political, or concern (over) cultural status” (Stewart 2010, 7). The concept of horizontal inequalities has also proven its utility in that is has been applied in numerous single and comparative case studies that persuasively demonstrate how the reality and perception of deliberate exclusion drives conflict (Marc et al. 2012)—for example, in Kenya, the Philippines, and Myanmar, and more broadly in Sub-Saharan Africa (Østby et al. 2009) and South and Southeast Asia (Parks et al. 2013). Horizontal inequalities play out not only at the group level but also at the regional level. They are particularly important in cases where regional and group levels overlap, specifically when excluded groups are geographically concentrated. Thus, Buhaug et al. (2011, 815) find “some evidence that regional inequality makes conflict more likely and that pockets of wealth in very poor states are especially likely to see civil conflict break out.” Similarly, Buhaug and Rød (2006, 327) present data that indicate that “dis- advantaged regions are generally more likely to host a secessionist rebellion.” Bakke and Wibbels (2006, 32) make the well-substantiated point that “in societies 14 Subnational Governance and Conflict characterized by regional ethnic concentration, . . . the likelihood of ethnic conflict rises as interregional inequality increases,” while Deiwiks et al. (2012, 298) note that “regional inequality is indeed robustly related to secessionist conflict onset.” These studies also confirm the exacerbating effect of politically excluding disadvan- taged regions from the center, which is poignantly expressed in the notion that the most motivated groups are those that hold grievances regarding their exclusion from political office and recent loss of status (Cederman et al. 2010). The importance of status loss as a driver of conflict is also confirmed by existing studies that systemati- cally consider conflict with subnational dimensions. Many scholars argue that lost autonomy makes for a particularly powerful grievance (Cederman et al. 2010; Gurr 1993; Gurr and Moore 1997; Siroky and Cuffe 2015). While motivation is clearly important to what drives conflict, it is not on its own sufficient to explain why conflict occurs. For conflict to begin, endure, and recur, it needs to be feasible. Beyond being motivated, conflict parties need to have the ability and capacity to engage in conflict. The factors most commonly identified as relevant in this context relate to demography, especially population diversity and size; geography, especially location, terrain, and neighborhood; and natural resources. In a sensitivity analysis of factors that determine civil war onset, Hegre and Sam- banis (2006, 531) confirm findings of earlier studies that had pointed out that “large population and low per capita income increase the risk of civil war” and that “civil wars are more likely to occur in countries with recent political instability and inconsistent democratic institutions; countries with small armies and rough terrain; countries located in war-prone, undemocratic regions; and countries with low rates of economic growth.” These factors map well onto the feasibility hypothesis of Collier et al. (2009, 2), who argue “that where a rebellion is feasible it will occur.” Such factors create conditions in which a political settlement is likely to break down because established institutional structures no longer satisfy groups capable of challenging an existing status quo. Demography is a key factor in determining a conflict party’s ability to mobilize around a specific identity, which is essential to induce collective action and overcome the perennial free-rider problem (Lujala et al. 2005; Siroky and Cuffe 2015). Yet feasi- bility is affected not only by population diversity but also that of population size. The existing literature agrees that larger groups find it easier to mobilize for collective action, even though size does not automatically translate into mobilization (Ceder- man et al. 2009). The risk for a violent outbreak further increases with significant changes in group power dynamics—for instance, through changes in a group’s size due to a specific population growth pattern (Walter 2009). This can be clearly demonstrated by demographic patterns in the Autonomous Region in Muslim Mindanao (ARMM) in the Philippines, which have changed dra- matically over the past one hundred years. In 1903, Muslims were approximately 80 percent of the population, and Christians were approximately 20 percent. But decades of internal migration have flipped the balance. Today across ARMM, approx- imately 80 percent of the population are Christians, and 20 percent are Muslims. Within these religious affiliations, approximately 13 percent identify as belonging to a range of indigenous peoples. However, while most Morosin the Southern Philip- pines, adhere to the faith of Islam, the ­ conflict in Mindanao is about the Subnational Governance and Conflict 15 marginalization of the Moro people, rather than about a conflict between Muslims and Christians in their religious identity. Territory has also been identified as a potentially enabling factor that can increase the feasibility of armed conflict. Several potential causal mechanisms explain this rela- tionship. Collier and Hoeffler (2005) argue that “rents are associated with large nontax income for the state or any other organization that can control the territory on which the rents are generated.” Consequently, resources increase the feasibility of conflict by providing the means for its conduct (i.e., resource finance conflict), as well as an opportunity for secessionists to mobilize support for their cause. “Secessionists not only claim ownership of the resources, they also claim that the national authorities are misusing the money—that it is being embezzled by distant elites” (Collier and Hoeffler 2005, 631–32). Territory and demography intersect in many ways, including—as illustrated by our case study of Mali—in large territories with low population density, which poses major issues for state institutions trying to project a positive presence. Buhaug et al. (2009) echo this finding in a broader context, noting that location in a general sense (absolute and relative distance from the capital, proximity to international borders, and avail- ability of lootable resources, especially gemstones and petroleum production) increases the likelihood of longer lasting conflicts. The likely explanation is that favorable loca- tions directly and indirectly increase both rebel fighting capacity and motivation. Closely related to location, another enabling factor that is widely considered critical for violent conflict onset, duration, and recurrence is a territory’s natural resources. In particular, lootable resources have been identified as a key driver of intrastate conflict. Early work by Collier and Hoeffler (1998, 2002, 2004) drew links between a country’s dependence on primary commodity exports and its risk of intrastate conflict. How- ever, Fearon (2005, 503) subsequently demonstrated that this association was “neither strong nor robust.” Instead, he argues that oil exports were a better risk indicator, not because of “better financing opportunities” (Fearon 2005, 503–4) but, because of the weaker state structures generally associated with oil-­ exporting states and the greater payoff for groups that seek either to capture the state as a whole or to secede from it with a significant share of its hydrocarbon wealth. Humphreys (2005) also presents evidence that weak state structures make violent conflict more likely but finds, con- trary to Fearon, that such state fragility is less likely linked to natural resource endow- ments, such as oil and diamonds, than to dependence on agricultural commodities. Other studies, too, have confirmed the critical importance of access to resources. For example, focusing on the question of lootable resources such as diamonds, Lujala et al. (2005) argue that “the incidence of rebellion is decided by motivation, opportu- nity and identity” and note that the same three factors also determine duration. Sug- gesting that diamonds only have a significant effect in intrastate conflicts that are fought along ethnic lines, Lujala et al. (2005) demonstrate that the underlying causal mechanisms for the onset of conflict relate primarily to the greed argument for moti- vation and the need to finance conflict as far as opportunity is concerned. This in turn determines the ability of such groups to continue fighting over a long period. Cru- cially, Lujala et al. (2005) make the point that this only holds for secondary (i.e., loot- able) resources, and only applies after production has begun, rather than upon the 16 Subnational Governance and Conflict discovery of diamond deposits. This finding is echoed by Humphreys (2005). There is some evidence that primary (such as non-lootable) resources decrease the risk of con- flict onset because revenues from their exploitation only strengthen the government (because of the investment and market access required). This puts the government in a position to better defend itself and/or buy off potential challengers. Duration Resources are also related to conflict duration. Humphreys (2005, 535) notes that “nat- ural resource conflicts are more likely to end quickly and are more likely to end with military victory for one side than with a negotiated settlement,” suggesting that rebel structures might be weaker and rebel groups more likely to be defeated when lootable resources are present. This implies a weak motivation to sustain mobilization, com- pared to more ideationally motivated conflicts (see also Weinstein 2005). It also implies that external actors interested in exploiting resources are more likely to intervene mili- tarily in support of one of the conflict parties, thus decisively tipping the balance of power on the battlefield. According to Walter (2009), power asymmetries also affect the duration of a conflict. In cases where the power between warring parties is not symmet- rical, conflict is particularly long and less likely to be terminated through a political settlement (Walter 2009). Long and intense conflicts often require different solutions than shorter, less intense ones (Balch-Lindsay et al. 2008; Collier et al. 2004; Cunningham 2006; Fearon 2004). This point is also made by Cederman et al. (2015), who find that “regional autonomy can serve a useful conflict-preventing purpose that keeps a group-government rela- tionship peaceful” if offered in a timely manner and combined with political inclusion (power sharing) at its center to increase the credibility of government commitments in the aftermath of violent conflict. The longer a conflict lasts, the more likely it is that its “original causes” transform and hence require different solutions. In this context, Weinstein (2005) argues that “resource-rich rebel groups are overwhelmed by opportunistic joiners, while those waging war with limited economic endowments attract committed soldiers to their movements.” This implies that in resource-rich contexts, grievance-motivated rebel- lions are more likely to be “crowded out” by greed-motivated ones. From a political settlements perspective, this is an important point: constraining feasibility might work to end a greed-­motivated civil war, but it is unlikely to create conditions for sustainable peace where conflict is motivated by grievance or where a grievance-motivated conflict is transformed into a greed-motivated one (Theuerkauf 2010). Put differently, conflict duration is likely to change power balances that must be reflected in a political settle- ment’s institutional design, but stability can only be achieved if underlying causes are sufficiently addressed as well. Otherwise, conflict is likely to recur. Recurrence The nature of conflict termination affects the likelihood of conflict recurrence. Empir- ical evidence suggests that conflicts terminated by a negotiated settlement are twice as likely to reignite as conflicts that end with a clear military victory (Toft 2010). More Subnational Governance and Conflict 17 specifically, Hartzell and Hoddie (2007) find that negotiated settlements are 34 percent less likely to result in renewed conflict, as compared to 58 percent for conflicts that are ended through a one-sided military victory. Therefore, recurrence is also related to flaws in the settlement, whether it was negotiated or externally imposed, and in its implementation (Chapman and Roeder 2007; DeRouen et al. 2010; Hartzell 1999; Mattes and Savun 2009; Pearson et al. 2006; Walter 1997). The content of any political settlement needs to match the context of the conflict, especially the prevailing balance of power between relevant groups. In other words, such arrangements must address not only the demands of conflict parties at the nego- tiation table but also the underlying drivers of which they are a direct or indirect expression. Settlements must also consider the need for credible commitments and sustainable implementation processes (Wolff 2011b). From Drivers to Mitigation One way to look at the numerous and often inconclusive findings on the causes of intrastate conflict is to distinguish between actionable and nonactionable causes. In other words, we need to better understand which causes can be tackled by policy inter- ventions, and which interventions can successfully address the complexity of drivers at work in most of today’s intrastate conflicts. In addition, we need to establish which contingent factors are likely to shape the success or failure of a conflict mitigation strategy that involves subnational governance arrangements. This can then provide a basis for designing effective interventions and for setting realistic expectations about their costs and consequences. From a political settlements perspective, this requires thoroughly understanding what drives those groups that have the capacity to destabi- lize a settlement and identifying the institutional structures that could accommodate their interests without marginalizing other powerful groups. Thus it is critical to look not only at challenger interests but also at the interests of those who defend an existing political settlement in order to articulate a feasible and viable institutional compro- mise that can underpin a new and sustainable political settlement. As noted above, there is a distinct strand in the literature that attributes causes of conflict to nonactionable factors such as mountainous terrain, distance from the state capital, natural resources, and disasters, including those increasingly related to climate change. These causes cannot be directly addressed but may in some cases be indirectly mitigated (Buhaug 2010; Buhaug et al. 2009; Buhaug and Rød 2006; Hendrix and Salehyan 2012; Koubi et al. 2012). By contrast, two other strands in the literature tie causes to greed (or opportunity), often related to natural resources (Ballentine and Sherman 2003; Berdal and Malone 2000; Collier 2001; Collier and Hoeffler 2004; Collier and Hoeffler 1998, 2005; Collier et al. 2009; Fearon 2005; Humphreys 2005; Ross 2004a, 2004b, 2006; Rustad and Bin- ningsbø 2012; Weidmann 2009; Wischnath and Buhaug 2014) and to grievances, espe- cially horizontal inequalities (Cederman et al. 2011; Deiwiks et al. 2012; Gurr 1993; Østby et al. 2009; Stewart 2010). Both types of causes can be more directly addressed by appropriate mitigation policies (Stewart et al. 2007). Yet another strand in the liter- ature distinguishes between motive, means, opportunity, and/or underlying and 18 Subnational Governance and Conflict proximate causes of intrastate conflicts as part of a more complex causal framework (Brown 2001; Cordell and Wolff 2009; Gurr and Moore 1997). This perspective offers additional opportunities to consider interventions as part of a crisis response— relatively short-term measures that limit access to means (e.g., border closures) or constraint opportunities (e.g., sanctions) and are therefore aimed at preventing a con- flict’s further escalation. Our review of the literature has thus far revealed that there is no consensus on what drives interstate conflict; what factors determine its onset, duration, and recurrence; or what shapes the motivation of individual and group actors to engage in intrastate con- flict. We can nonetheless draw a number of important conclusions to inform our sub- sequent reflections on subnational governance arrangements and their capacity to mitigate intrastate conflicts with subnational dimensions: 1. Not all drivers of intrastate conflict are likely to be equally responsive to mitigation by subnational governance arrangements. 2. Conflicts with subnational dimensions are likely to require subnational governance arrangements that are complemented by additional arrangements in order to estab- lish a sustainable new political settlement. 3. Prior to tackling a conflict’s root causes, it may be necessary to manage an acute crisis in order to create the conditions in which conflict parties and international development partners can constructively consider longer term, sustainable political settlements and how subnational governance arrangements function within them (i.e., there may need to be a transitional period as a new political settlement is nego- tiated and implemented). Subnational Governance Subnational governance spans all layers of governance, from the communal level to the central government, distributing varying degrees of formal and informal author- ity to communities, municipalities, districts, governorates, regions, states, and the center. The link between institutional capacity and political stability and security emphasizes the need to develop functioning subnational governance arrangements as part of a package of resilient and effective institutions and dispute resolution mechanisms. Governance Packages Governance arrangements are understood as the outcome of political settlements between elites (World Bank 2017). As is increasingly the case in fragile and conflict- affected states, a handful of political elites may control the central government and state resources and resort to violence to foster their political power and access to state resources (Justino 2017b). Effective governance interventions in fragile and conflict- affected settings have to take a multilayered approach and find entry points at national, regional, and local levels (Justino 2017a). Interventions must focus on the communal and municipal level if they are to build resilience (Marc et al. 2012), since this is the closest point of interaction between society and the state. Subnational Governance and Conflict 19 The existing literature recognizes that subnational governance arrangements are particularly effective when combined with other conflict mitigation mechanisms, especially forms of central power sharing that mitigate aspects of political exclusion (Cederman et al. 2015) and balance the potentially centrifugal implications of subna- tional governance arrangements (Weller and Wolff 2005; Wolff 2008). Horowitz (2007) suggests that a broad set of government institutions encouraging consensus building between different identity groups is more effective for reducing tensions. Echoing this finding, Hartzell and Hoddie (2007) argue that the more power-sharing arrangements are incorporated into peace agreements that underpin new political s ­ettlements— including what they refer to as territorial power sharing, or what we call subnational governance arrangements, the less likely it will be that conflict will renew. Subnational governance arrangements are also more effective when combined with other forms of governance. Tranchant (2016, 40) concludes that “the combination of regional autonomy and wide-ranging decentralization reduces incidence of civil wars by about 60–70 percent.” Similarly, Theuerkauf (2012) and Neudorfer et al. (2016) find robust empirical evidence that subnational governance arrangements on their own, and in combination with an electoral system based on proportional representation for the national legislature, are an effective conflict mitigation tool that reduces the risk of vio- lence in intrastate conflicts that have a subnational dimension. This dynamic also plays out at the subnational level. The political exclusion of groups within self-governing subnational entities tends to increase the likelihood of conflict (Diprose 2009; Diprose and Ukiwo 2008), whereas regional power-sharing arrange- ments generally have the opposite effect (Wolff 2004, 2011c). Related to the need to complement subnational governance arrangements with inclusive institutions that exist both regionally and nationally, there is a similar recognition that “to function well, autonomy frameworks require mechanisms for debating and resolving differences that arise between the autonomy and the state” (Woodman and Ghai 2013, 483). Forms of Subnational Governance Subnational governance as understood in this study represents a broad spectrum of governance arrangements, including political, fiscal, and administrative decentraliza- tion; regional autonomy; and federalism. This spectrum encompasses both arrange- ments of self-rule, which devolve significant political and governing authority to subnational units, and arrangements of shared rule, which provide subnational units access to decision-making processes at the center. Subnational governance building, reforming, and strengthening explicitly entail the (re)organization of a country’s gov- ernance structure at multiple levels. Broadly speaking, subnational governance is a way to transfer political, fiscal, and administrative powers as part of executive, legislative, and judicial competencies from the central government to subnational governance units. While there is no clear delineation between the various forms of subnational gover- nance, and while attempts to categorize forms of subnational governance are futile due to both their dynamic nature and the fact that most forms have overlapping character- istics, broad generalizations can be derived. Governance structures are commonly 20 Subnational Governance and Conflict classified as unitary, federal, or hybrid, but in some cases unitary states have a high degree of decentralization, and federal states may have a high degree of centralized control. However, comparative analyses suggest that most commonly unitary states with a high degree of decentralization grant greater autonomy at the municipal level, while federal states with a high degree of decentralization are defined by autonomy arrangements at the regional level (Weller and Nobbs 2010). Empirically, the most common forms of subnational governance arrangements employed for conflict mitigation are federations, federacies, devolutions, and local self-governance arrangements. A federation is characterized by “extensive self-rule with institutionalized shared rule,” usually in the form of “a constitutionally entrenched structure in which the entire territory of a state is divided into separate political units, all of which enjoy certain exclusive executive, legislative and judicial powers indepen- dent of the federal government” (Csergö et al. 2017, 496). Federacy arrangements involve “constitutionally entrenched extensive self-rule for specific entities.” Thus, “the main distinction between a federacy arrangement and a federation is that the former enjoys similar powers and constitutional protection as federal entities, but is distinct in that it does not necessitate territorial subdivisions across the entire state territory. In other words, federacy arrangements are a feature of otherwise unitary states” (Csergö et al. 2017, 496–97). Like a federation, a devolution can be applied to second-tier administrative units across the entire state or, like a federacy, only to selected territories in an otherwise unitary state. However, the degree of legal protection for entities with devolved powers is generally weaker and often extends only to protection by “regular” rather than con- stitutional laws (Wolff 2013b). Local self-government as a subnational governance arrangement for conflict miti- gation implies “the application of the principle of subsidiarity [and] involves the exer- cise of executive and administrative powers by lower levels of government, including regional and municipal authorities. It does not include legislative competences but provides the power to pass by-laws” (Csergö et al. 2017, 497). These conceptual distinctions are reflected empirically in our case studies, as illus- trated in Table 1. TABLE 1  Types of Subnational Governance Arrangements Bosnia and Herzegovina Highly decentralized governance structure Kenya Devolution Mali Unitary state system with local self-governance arrangements Myanmar Unitary state system with multiple federacy arrangements Philippines Unitary state system with a federacy arrangement Somalia Federation Yemen Federation (proposed) Subnational Governance and Conflict 21 Federations, federacies, devolutions, and local self-governance arrangements are not necessarily mutually exclusive. Local self-governance arrangements usually form part of federations, federacy arrangements, and devolved subnational governance set- ups. Moreover, these arrangements are further defined by varying degrees of decen- tralization, deconcentration, and delegation, whereby subnational governance units can enjoy varying degrees of political, administrative, or fiscal decentralization. Deconcentration denotes processes whereby regional offices of central ministries are established in local jurisdictions to guide policy decisions on the provision of local goods and services. Delegation is defined by a principle-agent relationship of the cen- tral government that makes subnational governments responsible for providing cer- tain local functions. However, the degree of decentralization is not always equal across the entire territory. For example, the governance model in the Philippines grants high levels of political decentralization through the election of local representatives, while fiscal arrangements are highly centralized; subnational governments depend on the central government for fiscal transfers. Nor do subnational governance arrangements need to be symmetrical in nature: within the same state, different entities can enjoy varying levels of autonomy. Such asymmetric federal or federacy arrangements allow for more flexibility than may be required to mitigate conflict in multiethnic societies that are prone to interethnic dis- putes (McGarry 2007; Swenden 2002; Weller and Nobbs 2010). Mali’s continuing instability can be partially explained by design flaws in the original system, which did not deliver on commitments made in the various peace accords. The North was offered special status in the National Pact, but this did not translate into a specific governance model that recognized the region’s cultural and socioeconomic particularities. Rather than choosing an asymmetrical arrangement in response to these demands and partic- ularities, the government opted for a unitary model with a high degree of symmetric decentralization. The complexity of a subnational governance arrangement can also pose a risk. This is a particular challenge in states with weak governance capacity and a lack of resources, such as those emerging from civil war. Tranchant (2016, 30) finds that the number of subnational governance units formed significantly affects peace and stability, and argues that “increasing the number of layers of governments may fuel conflict through the ensuing loss of policy coordination.” In some cases, the governance structure is defined by multiple asymmetric arrangements for different regions. Somalia represents a classic case in which various regions enjoy different degrees of autonomy: Puntland acquired the status of an autonomous federal state, while Somaliland announced its indepen- dence altogether. Subnational governance reform in fragile and conflict-affected states is often a dynamic and long-term process. Such is the case in Mali, where a political settlement gave subnational governance units the choice of either opting in or out of their auton- omy status (Weller and Nobbs 2010). Similarly, in the Philippines, the 2014 Compre- hensive Agreement on the Bangsamoro (CAB) included an “opt-in” clause, allowing the citizens of local government units outside the Bangsamoro core territories to peti- tion to conduct a vote to join the Bangsamoro. The intended territory for ARMM has 22 Subnational Governance and Conflict shrunk significantly since the first agreement (made in 1976)—it initially covered thir- teen provinces, whereas the Bangsamoro currently extends over the five provinces of ARMM and has the potential for some contiguous municipalities and barangays to join. Many of these communities have participated in previous rounds of referendums (in 1977, 1989, and 2001). Subnational governance in itself changes the power balance between the various parties, particularly between the center and the periphery, thereby changing the bargaining position of various actors. Thus, subnational governance arrangements are constantly renegotiated; the actual form of such arrangements remains fluid, reflecting the shifting nature of the political landscape. Subnational Governance to Mitigate Conflict A critical examination of existing knowledge on the merits and risks of subnational governance arrangements for conflict mitigation clearly reveals that such arrange- ments have genuine potential to reduce the risk of violent intrastate conflict by pro- moting the following: Inclusion • Granting minority and marginalized groups access to political decision making at the center and at subnational governance levels • Reducing the uncertainty and stakes of democratic processes • Empowering regional and local opposition parties to constrain the political power of dominant, national-level parties • Establishing checks and balances on a state’s authority at multiple layers of government • Reducing horizontal inequality • Promoting participatory and community-driven local governance • Creating space for political outsiders to enter local elections and to drive change and institutional reform Self-governance • Maintaining the state’s territorial integrity while granting minority and marginal- ized groups the autonomy to manage their own affairs • Bringing decisions closer to the people • Promoting policy flexibility • Responding to local needs and preferences Social cohesion • Addressing the grievances and protecting the needs of marginalized communities • Providing a nonviolent mechanism to resolve intercommunal tensions peacefully • Establishing and enforcing social norms within local communities Efficient governance • Building state presence in the periphery to extend institutional authority • Fostering the legitimacy of the state through strengthened state-society relations Subnational Governance and Conflict 23 • Improving local service provision and enhancing accountability for delivery • Facilitating the coordination and transmission of information among local networks Conflict Drivers Mitigated through Subnational Governance Political Exclusion Intrastate conflict driven by political exclusion can be mitigated through subnational governance arrangements that assign legislative and executive competences to subna- tional entities. Being able to elect and to hold accountable officials and representatives who make decisions directly relevant to different communities can increase a group’s sense of control over its situation. More “political” self-governance at the subnational level can also contribute to better integration at the center, for example, through regional representation (i.e., by establishing an upper chamber of regions or co-opting regional officials into central government bodies). Economic Exclusion Intrastate conflict driven by economic exclusion can be mitigated through subnational governance arrangements that assign competences and resources for regional eco- nomic development to subnational (legislative and/or executive) bodies. This can facilitate a more locally effective and accountable response to economic inequality. It can be undertaken by, for example, more deliberately and generously redistributing resources to poor regions or by assigning a fairer share of locally generated revenues to resource-rich regions. It can also feature centrally or externally funded (and locally administered) infrastructure development, including better integrating of remote or isolated regions into national transport, energy, and communications infrastructure. Social Exclusion Intrastate conflict driven by social exclusion can be mitigated through subnational governance by assigning the competences and necessary resources for delivering social services, such as those related to education, health care, and housing. Doing so can facilitate a more effective and efficient response to specific local needs. Subnational governance arrangements can help specific regions and their residents improve the quality and delivery of such services and “catch up” with more advanced regions, thus contributing to (a perception of) reduced horizontal inequalities. Cultural Exclusion Subnational governance arrangements can help address cultural exclusion by officially recognizing specific identity groups (defined by, for example, religion or language) and allowing such groups to administer policies critical to the preservation and devel- opment of their culture, independent from the interference of other cultural groups. 24 Subnational Governance and Conflict This can involve using specific regional languages, making special provisions for reli- gious worship, considering the importance of culture in the use of land and other resources, and acknowledging how these and other issues affect the implementation of (economic) development policies, the administration of justice, and/or the delivery of social services. Security It is particularly critical to address security concerns in the aftermath of violent con- flict. Depending on how far-reaching a concept of human security one wishes to adopt, subnational governance arrangements can increase groups’ sense of security by giving them more control over regional police forces, administration of justice, exploitation of resources, educational curricula, and more. In assessing how subnational governance arrangements can help mitigate conflict, it is important to bear in mind two caveats. First, the conflict-mitigating effects of sub- national governance arrangements are a function of their ability to address group/ regional grievances through a devolution of competences and resources that enables forms of governance that are more sensitive to local needs and that deliver improved results for otherwise separatist and/or insurgent groups. At the same time, these arrangements need to be carefully balanced by complementary governance mecha- nisms that preserve, and when needed enhance, the state’s viability and sustainability. Second, the adoption and implementation of subnational governance arrangements is neither cost nor consequence free, and these costs and consequences must be consid- ered and ameliorated. If these are not properly accounted for when a new political settlement is engineered, it is likely to fail. In other words, negotiating and implementing a new political settlement does not guarantee lasting political stability. This is in part because the process and outcome of negotiations—that is, what kind of institutional structures are established and how—may: 1. Not appropriately reflect society’s new balance of power; 2. Have unintended consequences that lead to further shifts in this balance of power; 3. Create different horizontal inequalities that produce new challengers to the political settlement; and 4. Establish institutions that simply do not work well on their own or in concert. This complexity cannot be fully elaborated here, but it can be illustrated with the following examples, prior to more in-depth case study–based investigations. In a broad sense, devolving political power to a subnational entity may empower a local majority to the detriment (e.g., political, economic, social, or cultural exclusion) of one or more local minorities. This may even be a deliberate policy to, for example, reverse the status of a long-suppressed group (e.g., a statewide minority in a previously more unitary system of government). An escalation of local tensions and possibly vio- lent conflict can be avoided through a number of measures, including: local power- sharing mechanisms to provide greater levels of political inclusion at the relevant Subnational Governance and Conflict 25 subnational level; personal (cultural) autonomy to allow different groups to self- govern regarding educational matters; the adoption of more than one official regional language; the passage and enforcement of national human and minority rights legisla- tion; and adherence to the rule of law within a unitary justice system, among others. Increasing subnational authorities’ control over financial resources—including giv- ing them the ability to raise their own independent revenue, decide on spending prior- ities, and/or vary the national corporate or income tax rate—may exacerbate regional inequalities, that is, make poorer regions poorer and richer regions richer. This increases the feasibility of and/or motivation for secession. Carefully tailored systems of fiscal decentralization that balance fiscal autonomy from, and control (redistribu- tion) by, the center can mitigate against such tendencies, as can national development institutions that work as independent government agencies or that adequately reflect, in their composition and decision-making procedures, the diverse range of regional interests. The proliferation of subnational governance institutions can cause bureaucratic overlap, and thus increase the risk of inefficiencies in service delivery. This can mani- fest itself through pressure on (limited) material, financial, and human resources. For example, the creation of a subnational layer of government between the central and local governments can trigger a skills drain from the local level up toward the regional. Likewise, it is conceivable that an influx might occur of poorly or inappropriately skilled bureaucrats from the center to occupy more prestigious positions at the regional level. While the former would weaken local governance, the latter would weaken regional governance and likely cause friction between both local and regional, and regional and central authorities. While there is no short-term solution, national and/or international investment in capacity building would be essential. Alternatively, or additionally, a more gradual and carefully sequenced devolution of competences, accompanied by appropriate resources, could be considered to enhance the positive, conflict-mitigating impact that subnational governance arrangements are meant to have. The implementation of subnational governance arrangements as a response to intrastate conflicts often involves complex institutional reforms, including, at times, the introduction of a whole new systems of government. This is partly a challenge to often limited state (financial and technical) capacity, which can be addressed through international assistance, as well as an issue that is likely to generate disputes within and between new or newly reformed institutions and their elites and bureaucracies. Such disputes can critically undermine peace processes by destroying nascent trust between former conflict parties, in the viability of negotiated agreements, and in their interna- tional sponsors. Consequently, it is essential to include policy coordination and dispute-resolution mechanisms into the broader institutional framework that comple- ments subnational governance arrangements—­ the former in order to avoid conflicts over policy making, the latter to resolve them constructively. This, too, will likely require international assistance during the negotiation of a new political settlement (to advise on which mechanisms will be useful) and its implementation and operation (by providing resources and expertise to build institutional capacity). 26 Subnational Governance and Conflict The engineered political settlements that our contributors’ case studies cover illus- trate this latter point well. Engineering new political settlements is not an easy task, as they need to properly reflect a new societal consensus. Even where they do, the context in which they are negotiated, implemented, and over time sustained is complex and often vulnerable to disruption and setbacks. Engineering new political settlements thus necessitates the cooperation of all relevant domestic actors that are part of the new political settlement with the international development partners that are critical to its success. Notes 10. According to the 2018 Global Report on Internal Displacement prepared by the Internal Dis- placement Monitoring Centre, as many as 11.8 million people have been newly displaced due to conflict, increasing from 6.9 million in 2016. Of these, almost 400,000 people were displaced in Somalia. 11. The 2018 Global Peace Index estimates that the global economic impact of violence and intrastate conflict has been the highest over the past 10 years, estimated at an equivalent of 12.4 percent of global GDP, or US$1,988 per person. The cost of US$14.8 trillion has been calculated in terms of purchasing power parity (Institute for Economics and Peace 2018). 12. United Nations General Assembly. 2017. Resolution Adopted by the General Assembly on 6 July 2017 “Work of the Statistical Commission Pertaining to the 2030 Agenda for Sustain- able Development” (A/RES/71/313)’. New York, NY. https://undocs.org/A/RES/71/313, 20–21. SDG target 16.6 reads, “Develop effective, accountable, and transparent institutions at all levels.” SDG 16.7 states, “Ensure responsive, inclusive, participatory, and representative decision making at all levels.” 13. Another term used in this context is incidence, which partly overlaps with onset (i.e., first instance) and duration. See, for example, Collier and Hoeffler (2002), Goldstone et al. (2010), and Koubi et al. (2014). 14. Such demands are particularly prone to conflict “not merely because of high motivation but also because separatism challenges the state to the highest possible degree” (Cederman et al. 2010, 106). Subnational Governance and Conflict 27 CHAPTER 3 Framing Subnational Governance to Secure Peace Applying subnational governance arrangements as a conflict mitigation mechanism is fundamentally about redesigning political power-sharing and power-dividing arrange- ments, and redistributing state resources. Generally, political power and state resources are consolidated at the center and dispersed according to subnational governance arrangements away from the center to the periphery. This is based on the assumption that the central government has the legitimacy and capacity to exert state authority to the periphery. In turn, broad acceptance of state authority and legitimacy rests on the pillars of strong state-society relations, which presumes state capacity to provide basic services at the local level. As a conflict mitigation mechanism, agreed-upon subnational governance arrange- ments are often entrenched in a political settlement’s legal framework. However, given the power shifts that result from redistributing political power and state resources, political settlements that feature subnational governance arrangements must address potential commitment problems among the conflict parties. A review of the literature reveals emerging themes that warrant a more detailed deliberation of the merits and risks of using subnational governance arrangements in the context of conflict and fragility. The following section elaborates on these themes. Subnational Governance to Share Political Power Subnational governance as a measure to distribute political power between central- and subnational-level governance institutions has become fundamental to most polit- ical settlements seeking to build peace. Subnational governance arrangements, which are sometimes also referred to as “territorial power sharing,” divide public policy func- tions between a central government and subnational authorities. However, such arrangements should be complemented by other institutional arrangements to func- tion effectively and to maximize their conflict-mitigating potential. Such complemen- tary mechanisms commonly include power sharing at the center to facilitate the representation and participation of otherwise marginalized segments of society in decision-making processes. Specifically, horizontal power-sharing arrangements at the center facilitate marginalized groups’ access to the central government and executive authority, while vertical power-dividing arrangements disperse political power to lower levels of government. 29 Power-sharing arrangements at the center can include representation in the national legislature, executive, and judicial institutions; in different branches of the security ser- vices and armed forces; in the civil service; and in a range of government agencies. They can also include specific rules of participatory decision making, such as the requirement to have qualified and/or concurrent majorities for a bill before parlia- ment for it to be able to become law; the parallel consent of a second chamber for leg- islation that affects subnational governance arrangements; and supermajorities to enact executive policy decisions. Power-sharing arrangements at the center can take different forms, such as a rotat- ing presidency, the appointment of a prime minister from the opposition party, or a (transitional) government of national unity. In cases where subnational governance arrangements complement central power sharing, members of a rotating presidency may be elected directly from subnational entities, or “joint tickets” may require an overall majority of votes, as well as a majority within a predetermined number of sub- national entities. Second chambers, in which subnational entities are represented, offer another mechanism through which subnational governance arrangements can be reflected in central power-sharing structures, as do rules for co-opting subnational office holders into central executive institutions. From this perspective, central power sharing offers subnational entities opportuni- ties to be represented and participate in national decision-making processes. In addi- tion, the complementarity of power sharing and subnational governance contributes to more viable state institutions overall because it ensures better policy coordination, enables dispute avoidance and resolution between different layers of authority, and gives subnational authorities a more tangible stake in national politics. In this way, it helps mitigate against the centrifugal tendencies often associated with subnational governance arrangements (Weller and Wolff 2005, Wolff 2011a). Therefore, central power-sharing and subnational governance arrangements are expected to reduce violent conflict by addressing the exclusion of marginalized groups and providing an alternate way to peacefully resolve disputes (World Bank 2017). They can also serve as a conflict mitigation measure by offering warring parties a compro- mise that reduces the stakes in the outcomes of democratic processes, such as presi- dential elections, and in the security architecture, such as by integrating forces from former adversaries. Establishing institutions that provide opposition parties or minority groups with an opportunity to participate in the political process is expected to reduce the likelihood of renewed intrastate conflict (Roeder and Rothchild 2005). According to Hartzell and Hoddie (2007), the more that power-sharing provisions across the four dimensions of state power (political, security, economic, and territorial— namely, subnational governance) are incorporated into political settlements, the greater the prospects for long-term stability. For instance, in Bosnia and Herzegovina, the 1995 Dayton Peace Accords featured a range of power-sharing arrangements, including provisions for central, territorial, and military power-sharing institutions and policies. In addition to the extensive subnational governance arrangements at the level of the state and the federation, the accords assign the country a power-sharing arrangement at the center through a bicameral legislative body and a rotating presi- dency that ensures representation for the three main ethnic groups. 30 Subnational Governance and Conflict Power-sharing arrangements may also be incorporated at the subnational level to organize politics within subnational entities, particularly where the delineation of subnational boundaries has created new local minority groups. Establishing forums for local decision making gives local communities a direct stake in the political arena and allows minority groups broader representation at a subnational level. Moreover, integrating quotas into a local governance mechanism, such as a local council, gives marginalized and vulnerable groups like women or the internally displaced an opportunity to participate where they have traditionally faced exclusion (Justino 2017a). As noted, power-sharing arrangements may be critical to initiating political settle- ment negotiations. Roeder and Rothchild (2005) recognize that power-sharing arrangements that may facilitate political transitions in the short term are an effective tool for ending violence in that they encourage the initiation of negotiations, even if they presumably fail to bring long-term peace to plural societies. Roeder and Roth- child (2005, 6) contest that “the very same institutions that provide an attractive basis to end a conflict in an ethnically divided country are likely to hinder the consolida- tion of peace and democracy over the long term.” In this context, power-sharing arrangements serve as a costly signal to former adversaries and thereby encourage adversaries to participate in peace talks. On the other hand, Martin (2013) argues that negotiated peace settlements that constitute territorial decentralization as a power- sharing mechanism are more likely to sustain peace than other forms of power shar- ing. Unlike central-level power-sharing arrangements that lack any genuine intention to cooperate, territorial power-dividing arrangements separate warring parties and preserve a degree of self-governance (Martin 2013). Dispersing power to different governance levels and branches empowers each party to exercise power to an extent that only affects a limited range of policies (Roeder and Rothchild 2005). It is important to bear in mind that during postconflict or postauthoritarian transi- tions, the conflict-mitigating potential of central power-sharing and subnational gov- ernance arrangements depends in significant part on the center’s commitment to implement decentralization as part of a new political settlement. Such a settlement must be based on a genuine commitment among conflict parties to avoid the recur- rence of war, rather than representing a means to consolidate a new regime’s political power, thereby exacerbating existing grievances or creating new ones. Aalen and Muriaas (2017) argue that the commitment of state authorities to implement political decentralization depends on the continuity or discontinuity of parties’ hold on the executive authorities. In postconflict settings, where the former opposition has gained control of the central government, implementing regularized subnational elections is more likely than in cases where the former regime has maintained its leadership posi- tion within the central government. In the first case, the former opposition that bene- fits from broad support among communities in the periphery views subnational elections as a mechanism to consolidate political power by mobilizing support and strengthening alliances with its established network (Aalen and Muriaas 2017). This problem has been apparent in Myanmar’s transition, especially in the results of the 2015 elections. The former opposition party, the National League for Democracy (NLD), achieved electoral dominance across the country, including in most ethnic Framing Subnational Governance to Secure Peace 31 areas. Amplified by the first-past-the-post electoral system, this means it had a major- ity or virtual majority in all but two states/regions (Shan and Rakhine). This signals a repeat of the earlier democratic period in the 1950s, when the presence of a single dominant national party left little room for minority ethnic representation. Combined with the fact that the president filled all 14 chief minister positions with members of parliament who represent the NLD (as expected), ethnic leaders have become con- vinced that the greater self-­governance they sought for decades can only be achieved by changing the constitution to feature a more genuine subnational governance sys- tem. In other words, Myanmar has thus far missed a distinct opportunity to institute a governance system that combines meaningful central power sharing with subna- tional governance arrangements that reflect its ethnic minorities’ long-established aspirations. Subnational governance arrangements may mitigate against violence by reducing the uncertainties of democratic processes and discouraging the potential domination of politics by majority groups or powerful minority groups. Such arrangements secure minority and marginalized groups’ access to central decision making, even if they have failed to gain a winning share in the electoral process (Roeder and Rothchild 2005). Thus, Hartzell and Hoddie (2007, 85) suggest that “power-sharing institutions and policies contribute to peaceful conflict management by appearing to reduce the risk that any single group will control all the levers of state power.” Power Sharing and Subnational Politics This points to another important factor concerning the design of complementary power sharing and subnational governance institutions, that is, the structure of the political party system. The issue here is the existence of regional and national parties, and the balance of power between them. The cumulative evidence that emerges from various studies has proven inconclusive and highly contingent. Brancati (2006, 657) noted that “large regions, many regional legislatures, upper houses of government elected or appointed by regional legislatures, and noncurrent national and regional elections, are factors that increase the strength of regional parties in countries.” This in turn is likely to “increase ethnic conflict and secessionism by reinforcing ethnic identi- ties, passing legislation that is harmful to regional minorities, and mobilizing groups to engage in ethnic conflict (Brancati 2006, 653)” because subnational governance pro- vides regional parties the resources to do so, a line of argument echoed by Chapman and Roeder (2007). On the other hand, Filippov et al. (2004) argued that decentralized national politi- cal parties (such as national political parties that have regional chapters and contest elections both nationally and regionally) were more likely to contribute to stable and integrated federations. A third type of argument emphasizes the importance of politi- cal inclusion of regional parties at the center (Bakke and Wibbels 2006; Cederman et al. 2015). If subnational governance arrangements are part of a broader institutional package that simultaneously creates incentives for integration at the center, they can contribute to sustainable peace even in the presence of strong regional parties (Weller and Wolff 2005). 32 Subnational Governance and Conflict CASE STUDY 1  Kenya—Reducing the Stakes of Winner-Takes-All Politics Over the past decade, Kenya has embarked on a comprehensive devolution program that represents an integral part of its 2010 Constitution. The turmoil of the 2007–08 postelection violence has given rise to widespread support for strengthened subnational governance arrangements. The Kenya National Dialogue and Reconciliation mediation process, which was launched in the wake of the postelection violence, has identified the need to address subnational governance in the form of power sharing and power dividing between different ethnic groups as a key pillar for a path toward stability and to address grievances and disputes between ethnic groups. Subnational governance is not a new concept to Kenya. The nation gained independence in 1963 with a constitutional framework that was based on subnational governance with eight semi-federal units. Known as the Majimbo system, it offered a high degree of regional autonomy and a compromise between what were, at the time, the two main political parties. The Kenya African Democratic Union represented a number of minority groups that expressed concerns about potential domination by a majority group and that preferred a federal system of government. Meanwhile, the politically dominant Kenya African National Union (KANU), which represented the two majority ethnic groups of the Kikuyu and Luo communities, who also predominantly resided in urban areas, advocated for a unitary system. Within a year of independence, the KANU dismantled the subnational governance system and replaced it with a highly centralized one. Although the Majimbo system was abandoned, the debate on subnational governance reemerged in the 1990s following the reintroduction of multiparty politics; it played an important role in the 2007 elections and the accompanying violence. The two main opposition parties campaigned on a platform of devolution, accusing the government of only serving a handful of communities based on their ethnic affiliations and excluding the rest of the country. Kenya has a plural society comprising over forty ethnic groups, of which the largest five account for over 60 percent of the population. The ethnic structure plays an important part in politics and is the main driver of intrastate conflict. Kenya has a majoritarian electoral system in which the “first-past-the-post” model incentivizes politicians to mobilize support along ethnic lines to outcompete their adversaries. Since no community represents a large enough share to dominate decisively, political elites form alliances along ethnic lines. Cycles of violence characterized by electoral politics and the marginalization of minority groups have been at the core of intrastate conflict in the postcolonial period, as political elites vie for control of the central government and exploit grievances over inequality and exclusion. Both the 1992 and 1997 elections were marred by state-sponsored ethnic violence in several parts of the country. Historically, the state has been the main engine for regional development and advancing community interests, particularly as communities without access to political power and state resources were excluded from state services. Regional boundaries also correspond with the geographic concentration of ethnic groups. Thus, these disparities generally implied “ethno-regional” disparities in (Box continued next page) Framing Subnational Governance to Secure Peace 33 CASE STUDY 1  Kenya—Reducing the Stakes of Winner-Takes-All Politics  (continued) development. Successive presidents allocated state resources to their home areas at the expense of other regions. The government of President Uhuru Kenyatta, a Kikuyu, and his deputy, a Kalenjin, who won the 2013 election, were seen as continuing this practice by favoring the two communities. Such practices make politics a high-stakes contest, since communities that lost in elections have had to relinquish their access to development funds and face socioeconomic and political marginalization. Unprecedented political violence engulfed the country after a dispute over the 2007 presidential election results. Pitted against each other were the ruling Party of National Unity (PNU), comprised mainly of the Kikuyu and allied groups, and the opposition party, the Orange Democratic Movement (ODM), comprised of minority groups and the Kalenjin, Luo, and Luhya. Since the dispute was fought along identity lines, the conflict split the country into two ethno-regional blocs and threatened Kenya’s existence as a nation-state. There was an emerging consensus that to prevent violence from recurring, a new political settlement was needed that would reshape the distribution of political power and state resources. However, the central government’s resistance to international mediation delayed initial efforts to reach a political settlement. The government finally acquiesced to pressure when the African Union stepped in. International mediation under the African Union Panel of Eminent African Personalities halted violence and initiated negotiations over a power-sharing arrangement. Subnational governance in the form of power-sharing arrangements sought to provide a framework that encouraged political elites to look beyond their narrow partisan interests and promote the country’s greater good. Disagreement over what the concept of power sharing entails for Kenya and how to operationalize it stalled the negotiation process. The PNU maintained that its candidate, Mwai Kibaki, was legitimately elected and that sharing power with the opposition was unconstitutional and would end multiparty democracy in Kenya. Meanwhile, the ODM demanded that the constitution be changed to create an office of the prime minister that had executive powers equal to the presidency. The proposed agreement stressed that the coalition government’s composition must be based on a principle of portfolio balance and reflect the parties’ parliamentary strength. An important element of the arrangement was that it created a prime minister position and two deputy positions. The prime minster would come from the party with the majority in parliament, and each party would nominate a candidate to serve as deputy prime minister. Both sides had an equal number of appointees in the cabinet of 50 seats for each party. However, efforts to accommodate marginalized groups also had implications on the effectiveness of the government, as this represents the largest government in Kenya’s history. The coalition government drafted and passed a new constitution that dispersed power both horizontally and vertically. Horizontally, power was redistributed among the executive, the legislature, and the judiciary. The constitution enhanced the (Box continued next page) 34 Subnational Governance and Conflict CASE STUDY 1  Kenya—Reducing the Stakes of Winner-Takes-All Politics  (continued) independence of the legislature and the judiciary. Furthermore, the constitution introduced a requirement for a winning presidential candidate to garner a majority of 50 percent-plus-one of the votes cast, as opposed to the previous mechanism stipulating that the candidate merely had to secure most of the votes. The candidate would also be required to garner a quarter of the votes in more than half of the counties, inferring a need to form broader coalitions. Vertically, power would be shared between the central government and subnational units, including a bicameral legislature. The constitution provides for a devolved governance structure based on a two- tier system: the central government and the 47 subnational units of the new county governments. The counties have power over issues such as agriculture, health care, and local infrastructure and have the right to raise revenue. Each county has a popularly elected governor and deputy governor, as well as members of the county assembly. The county assemblies have important powers of legislative oversight. For the most part, the model has served to eliminate Kenya’s winner-takes- all political dynamics, which were the main drivers of its conflict. Devolution contributed to the establishment of numerous positions at the subnational level and fostered inclusion among some marginalized groups. Thus, subnational governance arrangements sought to reduce the stakes of controlling the center and its resources by devolving both political power and state resources to lower levels of governance, ensuring greater access for various ethnic groups and in turn mitigating drivers of conflict. From an institutional and operational perspective, with the exception of certain areas, most functions have been devolved to subnational entities. The subnational governance arrangements were credited with contributing to the peaceful outcome of the 2013 elections. Indeed, it has been argued that subnational governance was the main factor encouraging opposition supporters to accept the disputed presidential polls. The opposition lost nationally but won in a number of counties, which contributed to the peaceful acceptance of election results. However, after the Supreme Court nullified the results of the August 2017 presidential elections— which pitted the incumbent president, Uhuru Kenyatta, against opposition leader Raila Odinga—due to irregularities, intercommunal violence accompanied a second runoff in October 2017. To thwart an escalation of the election dispute, Kenyatta and Odinga issued a joint statement in March 2018 expressing their commitment to promote dialogue and cohesion among rivaling ethnic communities. A positive outcome of devolution in Kenya has been the practice of “negotiated democracy” at the county level. Ahead of the 2013 elections, a major concern was the question of so-called cosmopolitan counties, where different ethnic communities reside within the same electoral district. A system of negotiated democracy, in which different communities agreed to share seats ahead of the polls, minimized the divisive nature of the democratic process. Arrangements were also made to allow minorities to be broadly represented in public offices at the subnational level. Such a formula was used in Migori County in western Kenya, (Box continued next page) Framing Subnational Governance to Secure Peace 35 CASE STUDY 1  Kenya—Reducing the Stakes of Winner-Takes-All Politics  (continued) for example, where elites from the Luo and Kuria communities agreed to share political power. The model was successfully replicated in other counties. A rising concern is that political elites’ commitment to devolution is dissipating. Initially, there was consensus that the system would be given 10 years before amendments could be introduced. Instead, the issue of devolution has once again become politically charged. Despite the fact that some members of parliament proposed drastically reducing the number of counties to lower costs, the idea has not gained much traction. Nevertheless, opposition parties and some governors have, on several occasions, claimed that powerful figures in government are determined to scuttle devolution in a repeat of the 1960s. Opposition parties and some governors are campaigning for a referendum, which they argue is necessary to safeguard devolution. Among their key demands is a constitutional amendment specifying that counties would collectively receive 45 percent of the national revenue. Parts of the country—such as the coastal regions and parts of western Kenya that have been marginalized by the national leadership—demanded higher fiscal transfers to the counties. Supporters of the proposal have dismissed the central government’s promise to increase allocations and insist that the increase should be clearly stipulated in the constitution, not left to the whims of the central government. Subnational Governance to Redistribute State Resources Subnational governance as a mechanism to redistribute state resources away from the center to the periphery and to share national wealth across regions is regarded as a viable strategy for reducing the likelihood of violent intrastate conflict fueled by eco- nomic grievances and disputes for control over national wealth. It is well known that regional inequality, economic marginalization, and grievances over equitable access to state resources are key drivers of conflict. Regional inequality significantly increases the likelihood of an escalation of intra- state conflict. Intrastate conflicts, specifically with subnational dimensions, are often fought over the form that national-wealth-sharing models take and the distribution of state resources between the center and periphery, and between subnational gover- nance units. In cases in which the territorial concentration of natural resources corre- sponds with the geographic delineation of identity groups, demands for greater regional autonomy or secession are more likely (Ross et al. 2012). Sambanis and Mila- novic (2014) analyze differences in regional units in countries with a decentralized governance system. They conclude that claims for autonomy are more likely to arise in wealthier regions, demonstrating that relative regional income is a predictor of demands for greater autonomy. Similarly, Jenne et al. (2007) find that economic dis- parities between regions and between minorities and majorities have an impact on the emergence of secessionist claims. Particularly throughout the 1990s, regional 36 Subnational Governance and Conflict inequality was linked to more extreme claims for self-determination. In such con- texts, economically marginalized minority groups leveraged separatist aspirations as a bargaining tool to reshape the distribution of state resources. It is well recognized that public finances are a central factor of intergovernmental systems and that the nature of the intergovernmental financial system shapes rela- tions between the central government and subnational entities (Mascagni 2016). Budget allocations and fiscal arrangements are informed not only by the goal of maximizing the efficiency of allocations but also by political questions, and in some instances by demographic characteristics, such as ethnic diversity. Similarly, wealth-sharing models for natural resource revenues are designed according to political incentives and subnational actors’ bargaining power (Mascagni 2016). In Somalia, the semiautonomous region of Puntland enjoys significant fiscal auton- omy, and regional authorities are authorized to collect revenues. Based on the bottom-up arrangements for subnational taxes, Puntland is authorized to collect a range of taxes locally that are normally assigned to the central government (World Bank 2014). As is the case for other governance models, there is no consensus on the merits of subnational governance arrangements in the form of fiscal autonomy or wealth shar- ing as a conflict mitigation mechanism. In an evaluation of the linkages between fiscal decentralization and intrastate conflict in 77 countries between 1972 and 2000, Ezcurra (2015) concluded that devolving fiscal power to subnational governance levels effec- tively reduced the incidence of intrastate conflict. On the contrary, Binningsbø and Rustad (2012) examined the potential of wealth-sharing arrangements in fostering peace in postconflict settings and found that such arrangements have not been proven to successfully sustain peace, whereas inadequate designs and a failure to implement have been identified as central obstacles to success. However, at least in the short term, subnational governance arrangements with financial dimensions have a positive impact on the resolution of intrastate conflict by initiating dialogue between adversaries. The transfer of state resources or the granting of fiscal autonomy can serve as a signal to, and foster confidence among, rebel groups and marginalized communities to win buy-in for a political settlement. Offering con- cessions to opposing factions in the form of fiscal autonomy or intergovernmental transfers provides moderate leaders with the internal leverage necessary to appease hard-liners and to generate support from their constituencies (World Bank 2014). For instance, in Bosnia and Herzegovina, the complete separation of fiscal responsibility according to principles of full subsidiarity was intended to calm the fears of former adversaries and induce their support for the political settlement. The central govern- ment’s fiscal management responsibilities are limited. The expenditures are con- strained by a fixed amount of indirect taxation revenues that is set aside for the central government before allocations to entities are determined. For fiscal decentralization to serve as an effective mechanism for conflict mitigation through its capacity to strengthen the social contract by increasing accountability of public officials, constituencies must regard their local government as the tax-raising authority. This will in turn put pressure on local government officials to deliver ser- vices and make them more accountable to the local community (Mascagni 2016). Framing Subnational Governance to Secure Peace 37 However, in fragile and conflict-affected countries, any tax-raising scheme must account for the fact that a significant share of economic activities are informal and often illicit. In the context of a flourishing shadow economy, it is also likely that local communities already carry the burden of informal revenue-collection schemes, such as protection rackets or other forms of coercive extraction from non-state actors. This adds to the fact that in conflict settings, local constituencies often lack the resources to cover their basic needs, let alone to pay taxes. Sharing the Benefits of Natural Resources Questions regarding wealth sharing are at the heart of many intrastate conflicts in countries that depend on natural resource revenues as their main source of public income. Conflict in this context can emerge in many forms, including through politi- cal competition for control of the central government, through attempts to secure leadership positions in central government institutions that manage national revenues, or through regional conflict, with oil-producing regions promoting separatism and marginalized regions demanding a fair share of the national wealth. In addition to fueling conflict for control over resources, both central and subnational government entities may face political unrest as constituencies express grievances over how national wealth is used (Mascagni 2016). In Yemen, for example, oil fields are mainly located in two regions in the South (Shabwah and Hadhra-mout) and to a very small extent in the North (in Mareb). Mareb also produced the country’s liquid natural gas for local consumption and exports. The two oil-producing regions in the South host only about 7 percent of the population. The 2015 amendments to the 1991 Constitution foresaw a central-level consolidation of oil revenues to be added to the general pool of public revenues and redistributed through transfers to subnational governance levels. In the absence of asymmetric fiscal autonomy, the oil-producing regions do not benefit disproportion- ately over the other regions. However, failing to design arrangements that benefit oil-producing regions or at least compensate for negative externalities associated with natural resource extraction has the potential to cause grievances among local commu- nities and drive social unrest and violent conflict. Various forms of subnational governance have the potential to address intrastate conflict driven by the presence of natural resources and disputes over the distribution of related revenues. Subnational governance arrangements that address the redistribu- tion of state resources can be categorized as vertical (between the center and the regions) or horizontal (between the regions). These include authorizing the central government to collect revenues and set budget allocations, transferring a share of rev- enue to local governments, or granting subnational governance entities the authority to collect taxes. Of these, assigning the central government the authority to determine fiscal policy is only feasible if subnational governance entities have the opportunity to participate in budget decision making, since postconflict settings are laden with dis- trust between former adversaries (Ross et al. 2012). 38 Subnational Governance and Conflict Ensuring the Provision of Adequate Financial Resources Subnational governance arrangements must be underpinned with the resources neces- sary to enable their respective units to function and assume their competences and responsibilities (Narang Suri 2016). Failing to account for a political settlement’s financial requirements may cause violent conflict to recur, as commitments made in political settlements cannot be implemented, which in turn fuels distrust and resent- ment among disenfranchised groups. Recent findings suggest that “the commitment to devolving resources and autonomy from distant central governments to local gov- ernments may increase the legitimacy of the local government, which in turn trans- lates into more cooperative behavior and stronger cohesion among groups” (World Bank 2017, 122). On the other hand, limiting a subnational governance entity’s fiscal autonomy is one way a central government can maintain control despite granting political and administrative autonomy. In fragile and conflict-affected states, fiscal dependency of subnational governance units on the center is high. According to the 2012 World Bank Public Expenditure Tracking and Facilities Survey, 86 percent of fragile and conflict- affected states ranked in the lowest quartile for budget autonomy (UNDP 2016). Esti- mates suggest that 70 percent of local governments in such settings depend on fiscal transfers from the central government, compared to the global average of 52 percent (UNDP 2016). Depending on financial resources (such as grants) from the central government undermines a subnational unit’s degree of self-governance. The implementation of subnational governance arrangements is enormously expensive and requires signifi- cant state funds, particularly if new government entities are formed and the institu- tional capacity of subnational administrations is weak or nonexistent. Economic constraints remain a key obstacle to implementing a series of peace settlements in Mali. The central government faces serious challenges in extending authority and delivering services in low population density areas. The three north- ern regions of Gao, Kidal, and Timbuktu cover three-quarters of the country’s territory but have only about 9 percent of the population. In 2013, average national population density was 13.5 persons per square kilometer, as compared to 0.29 per- sons per square kilometer in Kidal. This makes all forms of connectivity more challenging and expensive. According to Mali’s local governance regulations, local revenue collection was assigned as the primary source of income for local govern- ments. Due to the low population density, the local tax base is insufficient to sup- port the higher cost of delivery; thus local governments will continue to rely on intergovernmental transfers. The per capita cost of maintaining three levels of subnational governance is much higher in the North than in the rest of the terri- tory. Budget allocations are based on a set of equalization indicators, including local revenue capacity, remoteness, and population size, which seems to dispro- portionately benefit—though not always ­positively—the North. In 2013, for exam- ple, Kidal had a budgeted allocation that was almost four times that of any other region. This generates a sense of resentment among the southern regions. How- ever, overall transfers remain extremely limited and insufficient to ensure that the proposed subnational governance arrangements function properly. Framing Subnational Governance to Secure Peace 39 There is an inherent trade-off for the central government in seeking to grant signif- icant fiscal autonomy or to transfer large shares of public revenue to subnational enti- ties in order to appease opposing parties. At the same time the government needs to secure access to sufficient financial resources to finance the recovery process and the basic services delivery that are paramount to providing quick peace dividends. Fragile and conflict-affected countries are particularly cash stripped and often face looming economic crises (World Bank 2014). CASE STUDY 2  The Philippines—Exploring the Merits of Accommodation through Asymmetrical Fiscal Autonomy In the Philippines, asymmetrical fiscal autonomy seeks to accommodate demands for self-governance made by separatist groups of the southern region of Mindanao. Political groups and armed factions of the southern provinces, which have a Muslim Bangsamoro majority, have called for secession since the country’s independence in 1946. Subnational governance arrangements, foreseen in a recent political settlement between the central government and the main opposition group, aimed to satisfy Moro political movements’ long-standing aspirations for greater autonomy, while correcting years of actual and perceived neglect of the region’s development needs. The failed implementation of the 1996 Final Peace Agreement between the central government and the Moro National Liberation Front (MNLF) led to a rekindling of tensions followed by renewed reconciliation efforts in the form of the Bangsamoro peace process. Despite the main parties’ efforts to advance the implementation of the Bangsamoro peace agreement, genuine autonomy—particularly fiscal autonomy—has not been realized in any of the political settlements to date. Historically, the Philippines has had a highly centralized governance structure. Subnational governance first emerged as a central issue in the country’s political process in the 1950s. Since 1971, the Philippine government experienced a rise in armed opposition from Moro political and military groups calling for greater autonomy for the island of Mindanao. Frustrations fueled by the Moro population’s growing sense of injustice escalated into armed conflict between the MNLF and the Armed Forces of the Philippines in 1973. In December 1976, the government and the MNLF agreed in principle to the creation of an autonomous region in the southern Philippines. However, relations between the MNLF and the Ferdinand Marcos regime deteriorated, and the armed struggle continued. Almost every national administration since 1987 has proposed some form of asymmetric autonomy as a mechanism to resolve the long-running conflict. Upon taking office in 1986, President Corazon Aquino revived peace talks with the MNLF and secured the inclusion of provisions that foresaw the creation of the Autonomous Region in Muslim Mindanao (ARMM) in the 1987 Constitution. Over a decade after the adoption of the constitution, President Gloria Arroyo supported a September 2001 referendum ratifying the amended and expanded Organic Act of the ARMM, which further defined the regional administration’s powers and responsibilities. (Box continued next page) 40 Subnational Governance and Conflict CASE STUDY 2  The Philippines—Exploring the Merits of Accommodation through Asymmetrical Fiscal Autonomy  (continued) Meanwhile, a breakaway group of the MNLF—the Moro Islamic Liberation Front (MILF) —formed after the 1976 agreement, continued the armed struggle through the 1990s. Formal negotiations were opened between the government and the MILF in 1997. In 2000, clashes between the military and the MILF escalated dramatically and displaced hundreds of thousands of people, prompting President Arroyo to restart peace talks with the MILF in 2001. Against the background of these consecutively failed negotiations, peace talks were revived after President Benigno Aquino III was elected in 2010. After the Framework Agreement on the Bangsamoro was signed in 2012, the government and the MILF signed the Comprehensive Agreement on Bangsamoro (CAB) in 2014, which draws together all interim agreements since 1997. In return for the MILF abandoning its demands for independence, the CAB provides “recognition of the justness and legitimacy of the cause of the Bangsamoro people and their aspiration to chart their political future through a democratic process that will secure their identity and posterity and allow for meaningful self-governance.”a The core element of the CAB is the agreement to create a new autonomous political entity that was granted a large degree of political and fiscal autonomy. The CAB included a detailed annex on revenue generation and wealth sharing, regulating the autonomous fiscal administration of the Bangsamoro and seeking to end the region’s dependency on fiscal transfers from the central government. While the peace agreement remains in force, and a number of the joint transition bodies it created continue working on the formal steps agreed on in the CAB, key elements of the transition to the Bangsamoro are on hold. Despite positive indications after the CAB was signed, the draft legislation adopting the settlement provisions into the national legal framework ultimately stalled in Congress. Each time peace appeared to be at hand, different arms of the state intervened to reset the negotiations—the executive in 2000, the judiciary in 2008, and the legislature in 2015. As one of his main campaign promises, President Rodrigo Duterte, who took office in 2016, advocated support of the agreement and for broader federalism, pushing the legislature to pass what was needed to formalize the CAB. The Bangsamoro Organic Law formally establishing the Bangsamoro Autonomous Region to replace ARMM was signed into law in July 2018, and resoundingly endorsed in a plebiscite on January 21, 2019. The Bangsamoro Transition Plan, which will guide the implementation of the changeover from the ARMM to the BARMM over a three-year period until regional elections in 2022, was approved by the (interim) Bangsamoro Transition Authority in June 2019. The country also has a long history of economic marginalization overlapping with social and political inequalities. Communities living in regions that are home (Box continued next page) Framing Subnational Governance to Secure Peace 41 CASE STUDY 2  The Philippines—Exploring the Merits of Accommodation through Asymmetrical Fiscal Autonomy  (continued) to a majority of the country’s Muslims continue to suffer from higher poverty rates and lower human development indicators than the rest of the country. Human development indicators from 2005 suggest that out of the five poorest provinces in the Philippines, four are in southern Mindanao, and the ten poorest provinces are also the nation’s most fragile and conflict-affected territories. Political settlements that were intended to grant greater political and fiscal autonomy and promote regional development to address historical regional inequalities have been inadequately implemented. Many key elements of the 1996 Agreement were implemented, including the establishment of the Special Zone for Peace and Development, but the MNLF remained critical of the inability of various arms of the state to fulfil the commitments outlined in that agreement. ARMM was assigned competences in 1989 and 2001, periods that were characterized by a nationwide devolution of power from the center to local governments. The 1991 Local Government Code was the most sweeping law—it changed the country’s intergovernmental system and fiscal rules, dramatically increasing the funds allocated to subnational governance units. Yet local governments in the Bangsamoro Autonomous Region continue to depend on national government financing for the majority of their needs, with around 66 percent of total revenue coming in the form of nationally determined allocations, where national-level state officials make decisions on intergovernmental transfers to the Bangsamoro Autonomous Region. The government of the Bangsamoro Autonomous Region was granted “fiscal autonomy,” but only with respect to the local funds it could generate or the funds it was awarded from the national government. Crucially, the budgets for the provinces, cities, municipalities, and barangays that fall within the Bangsamoro Autonomous Region remained unchanged. The vast majority of funds for these local governments’ operations come from national allocations, and therefore the degree of influence or control of the government of the Bangsamoro Autonomous Region over these subnational governance units is extremely limited. Provinces, cities, municipalities, and barangays are preexisting entities with their own elected local governments, and as such guaranteed resources under the national decentralization provisions of the 1991 Local Government Code, making them effectively “autonomous” from ARMM control. This situation has complicated the regional government’s efforts to assert its autonomy, control its budget, and manage corruption. In this sense, ARMM has been doubly trapped between a truncated devolution process from above and a poorly integrated system of local government units from below. Incoherent fiscal decentralization has been central to ARMM’s failure to perform effectively as an autonomous region. With regard to fiscal autonomy, the CAB provided for an annual block grant to the Bangsamoro (subsequently proposed at 4 percent of the national government’s share of total revenue collection, which was designed to represent the proportion of ARMM population within the nation), and also included provisions for a Special Development Fund to finance “rehabilitation and development” programs. The (Box continued next page) 42 Subnational Governance and Conflict CASE STUDY 2  The Philippines—Exploring the Merits of Accommodation through Asymmetrical Fiscal Autonomy  (continued) Bangsamoro Organic Law foresees a range of fiscal measures—including the transfer of an annual block grant from the central government and an increase in the share of public revenues that have been raised in the Bangsamoro provinces— to be returned to the region to reach 75 percent. The central government has increased funding for ARMM in order to correct historic imbalances and support the peace process. Both the regional budget and national agency budgets intended for the region have grown dramatically. The combined allocations to ARMM are estimated to have grown by 37.9 percent in 2011 and 72.4 percent in 2012. However, some fundamental constraints to the ARMM government remain, including low regional revenues, extreme dependence on the national government, limited control over the total budget allocations for ARMM areas, and limited program and financial management capacity. When combined with vulnerabilities in budget execution and accountability, the situation impedes service delivery and threatens development outcomes. The most significant shift in the CAB is the introduction of a ministerial form of government that operates within the presidential unitary system of the Philippines. As the only nominally autonomous region among the nation’s 17 administrative regions, ARMM is unique in having a regional government headed by a regional governor and its own regional legislative assembly, which sought to create a balance of power within the Bangsamoro region. An open democratic competition for power would also create opportunities for groups other than the MILF to exercise regional power, whereas the existing systems vest almost all regional powers in the office of the governor. In terms of conflict mitigation, evidence from the creation of ARMM and the Bangsamoro Autonomous Region is mixed. Since ARMM’s formation in 1991, intrastate conflict has alternately intensified and waned, but the powers allocated to the region have not substantially changed. Viewed through the prism of conflict mitigation and the sustainability of political settlements, it is apparent that the various attempts to craft “autonomy” have had significant deficiencies and are perceived as having failed to provide the region with the requisite fiscal autonomy. There is more emphasis on Bangsamoro’s fiscal autonomy, both through defined revenue-sharing arrangements and the creation of intergovernmental fiscal institutions; however, fiscal autonomy may require more, rather than less, coordination between levels of government, which may bring the central and regional government closer instead of increasing the degree of separation between the two entities. Moreover, fiscal autonomy is merely one element in a spectrum of demands necessary for the Bangsamoro to meet their aspirations. Effectiveness, efficiency, and transparency of the new Bangsamoro authorities will be equally as important as fiscal autonomy, particularly in a context in which the conflict between the MILF and the central government remains just one of multiple ongoing sources of violence. Additionally, the Bangsamoro Autonomous Region is facing high levels of localized violence, including terrorist attacks committed by violent extremist groups (Box continued next page) Framing Subnational Governance to Secure Peace 43 CASE STUDY 2  The Philippines—Exploring the Merits of Accommodation through Asymmetrical Fiscal Autonomy  (continued) such as affiliates of the Islamic State in Iraq and the Levant (ISIL). To conclude, the drivers of conflict at the national and regional levels are highly complex, and fiscal autonomy has to be combined with a set of governance arrangements to achieve a resolution to the conflict that lasts. Note: a. The Comprehensive Agreement on the Bangsamoro, p. 1: https://peacemaker.un.org/sites/peacemaker .un.org/files/PH_140327_ComprehensiveAgreementBangsamoro.pdf Subnational Governance to Extend State Authority In cases in which central state institutions have almost nonexistent capacity, how can political power, financial resources, or administrative responsibilities be devolved to the periphery? In some instances, non-state armed groups temporarily hold terri- tory that is not under credible state authority (United Nations and World Bank 2018). For example, in 2012, a Tuareg separatist group took control of large swaths of northern Mali before central government forces backed by French soldiers suc- ceeded in regaining state authority. In some states, large parts of territories remain outside the control of the formal government. Control over territory and the extension of state authority to the periphery is a prerequisite for effective subnational governance. State sovereignty implies that the state holds authority across the entire territory. The establishment of what Felbab- Brown et al. (2018, 120) call “alternative local orders” is the result of a state’s weakness and inability “to provide a full spectrum of governance over the entire national terri- tory ostensibly under its authority.” Fragmented state authority and weak governance capacity are common in fragile and conflict-affected states. The central government’s failure to extend state authority to the periphery and assert its monopoly over the legitimate use of force results in a security vacuum that non-state actors can exploit. Non-state actors that might assert political power over local communities include traditional authorities such as tribal leaders, religious figures, and civil society organi- zations (Marc et al. 2012). In the absence of state authority and the rule of law, informal mechanisms emerge to regulate intercommunal relations and to exercise governance functions. There is a misconception that the collapse of the formal state represents the end of governance. Rather, a process unfolds that encourages the formation of informal governance sys- tems (Justino 2017a). In fragile and conflict-affected settings, there is a major gap between de jure and de facto authority held by various actors. Whereas the first is assigned by the constitutional framework and often underpinned by electoral legiti- macy, the latter is acquired through the threat or use of violence (Acemoglu and Robinson 2006), or through traditional authority. Efforts to extend state authority by strengthening subnational governance must take into account the interplay between informal and formal governance mechanisms. 44 Subnational Governance and Conflict The process of forming a state and establishing state authority through subna- tional governance reform is dynamic and driven by interactions between various state and non-state actors. International development actors engaged in subnational governance reform have to balance the need for broad inclusion with the risk of empowering spoilers. Failing to include non-state actors that have held territory and developed alternative governance systems in the postconflict political settlement and the subnational governance reform process may result in renewed conflict and create an environment of “no peace, no war” (World Bank 2017, 131). On the other hand, enhancing the role of elites—whether they are affiliated with political factions or are leaders of armed groups—in the subnational governance reform process risks empowering these actors and thereby exacerbating local communities’ vulnerability and marginalization (Gaynor 2016), as well as creating, and then rewarding, spoilers of the peace process. Some nondemocratic governments use subnational governance as a way to pene- trate ungoverned territories. This is undertaken by transferring central government officials to subnational entities to take up positions in local governments. Jackson (2016, 750) argues that “decentralization has been politically subverted by elites who use it to project their own power through political manipulation of local political sys- tems or as an extension of political control.” In such contexts, interventions to build and strengthen subnational governance may have detrimental effects on a country’s stability. Strengthening subnational governance as a measure to extend state authority in a context where the central government is authoritarian or its democracy remains flawed risks reinforcing and extending the central government’s repressive capacity, and likely disproportionately affects minority groups. Therefore, the presence of polit- ical elites who co-opt local leaders has also been identified as a driver of violent con- flict and instability (Justino 2017a; Narang Suri 2016). Where the state fails to extend authority and to effectively exercise subnational gov- ernance functions, political power and the monopoly on violence are held by a multi- tude of state and non-state actors, to varying degrees. As Marc et al. (2012, 27) argue, in fragile and conflict-affected settings, “state authority is often fragmented and con- tested by a variety of actors in society.” Ungoverned spaces in the periphery are condu- cive to localized conflicts, as armed non-state actors attempt to reshape the distribution of political power and “bargain” over the monopoly of violence within subnational territory (World Bank 2017). Thus, the presence of non-state actors raises the level of violent conflict as these actors fight for control over territory. Felbab-Brown et al. (2018) echo the finding that competition between non-state actors drives local dis- putes and communal violence. In rare cases, the prevalence of conflict, fragility, and violence can result in the col- lapse of both formal and informal authority (Marc et al. 2012). If the state fails to rapidly reassert control in a postconflict setting, war economies and illicit activities flourish in ungoverned spaces (UNDP 2016). Following the seizure of peripheral terri- tory, the central government must have the institutional and political capacity to fill ungoverned spaces. To permanently regain control, the central state has to build legit- imacy within local communities. The goal of extending state authority across the entire territory should be to strengthen those state-society relations that will serve as the foundation for a sustainable social contract. Framing Subnational Governance to Secure Peace 45 CASE STUDY 3  Somalia—State Collapse and Loss of Monopoly on the Use of Force Somalia has experienced the longest-running period of complete state collapse of any country since the postcolonial era. Today, Somalia’s federal government has reestablished its authority and is navigating the very challenging process of rebuilding political power, gaining monopoly on the legitimate use of force, and extending state authority across the country. The central government’s continued weak position in this process stands in contrast to the self-declared independent region of Somaliland in the North, which has largely enjoyed political stability, security, and the ability to build institutional capacity. Somalia’s decades-long civil war was defined by interfactional fighting, inconclusive foreign interventions, and failing peace processes, which left the central government without territorial control. Access to and distribution of political power and state resources at the national and regional levels have been at the heart of disputes between political elites, rival clans, and local communities. At the core of the conflict was a continued lack of consensus over Somalia’s governance model, its intergovernmental and state-society relations, and the role of Islam in society. While Somalia’s society is homogeneous in terms of religious and ethnic composition, the conflict was marked by a high degree of fragmentation. The conflict was mainly fought along clan lines. Subsequent support for, or opposition to, federalism or decentralized authority among Somalia’s elite and major clans fluctuated throughout the conflict, largely dependent on the political and economic gains they could expect from the respective form of subnational governance. This demonstrates the inherent political nature of subnational governance reforms and that preferences of national actors are informed by vested interests rather than considerations of governance efficiency or conflict mitigation. The federalist agenda was first put on the table during the 2002–04 Mbagathi peace process, but progress was only possible following the adoption of the 2012 Provisional Constitution and the inauguration of the first permanent federal government of Somalia. The provisional constitution proposed a federal governance system be formed on the basis of the country’s 18 historical regions; it broadly outlined that the central government and the regions share political power and government competences. The constitution also established power-sharing arrangements in the national legislature, where representatives from the federal states form the upper chamber of parliament. Detailed regulations regarding the separation of powers and competencies were expected to be outlined in the final constitution, which was scheduled to be finalized before the 2016 elections. However, despite commitments pledged by the government and by regional leaders, the constitutional review process is still ongoing, albeit with high levels of delay and contest. As a result of the internal power struggles, with different interests vying for political power and access to state resources, many of the elements outlined in the provisional constitution have not yet been implemented. Historically, Somalia’s process of strengthening subnational governance has been accompanied by violent disputes over territory and stalled by the central government’s lack of legitimacy, authority, and institutional capacity. (Box continued next page) 46 Subnational Governance and Conflict CASE STUDY 3  Somalia—State Collapse and Loss of Monopoly on the Use of Force   (continued) Since the collapse of the Barre regime in 1991, no central government has been able to consolidate state presence outside the capital or regain the state’s monopoly on the legitimate use of force. Due to persistent insecurity, elections scheduled for October 2016 had to be postponed. Instead of the envisaged universal elections, an indirect electoral process was pursued. Clan elders nominated 14,000 electoral delegates, who in turn appointed the 275 members of the Lower House of Parliament, while regional assemblies elected the members of the Upper House. The new Parliament was inaugurated in December 2016; it nominated Mohamed Abdullahi “Farmajo” Mohamed to serve as president. In June 2018, the National Security Council, a forum of political leaders of the federal government and the federal member states, reached a preliminary agreement on an electoral model for multiparty elections. The provisional constitution provides for a power-sharing arrangement through the position of a prime minister. However, former President Hassan Sheikh Mohamud, dismissed three cabinets and three prime ministers within the first three years of his administration. As a response, in 2015, parliamentarians threatened to launch an impeachment process against President Mohamud, claiming he had overstepped his constitutional powers. The continuous change in leadership and cabinet positions further undermined the central government’s ability to function. Questions regarding resource sharing continue to represent a main point of contention between central and regional authorities, as leaders of the federal member states complain of insufficient fiscal transfers. The state-building process has been further complicated by the varying statuses of the federal states in relation to the central government. Somaliland declared independence in 1991, and although its status has not been internationally recognized, the region has enjoyed relative peace since. In July 2018, the newly elected president reiterated his call for talks between the federal government and Somaliland. Puntland meanwhile announced its status as an autonomous region in 1998 but has since expressed its intentions to remain a semiautonomous federal entity within the overarching Somali state. Four new federal member states have been formed, including the Interim Juba Administration, the Interim South West Administration, the Interim Galmudug Administration, and the Interim HirShabelle Administration. Some regions are engaged in ongoing negotiations over the creation of federal states, while the aspirations of other regions to merge and form a federal state have been hindered by competing territorial claims. The provisional constitution stipulates that the federal government is tasked with establishing the boundaries and defining the territory of federal entities, but adds that two or more regions can join to form a federal member state. Somaliland and Puntland are in dispute—having intensified military clashes since early 2018— over the regions of Sool and Sanaag, which in turn are considering whether to form their own separate, federal state. Most recently, in May 2018, Puntland launched an offensive to reclaim territory of the Sool region from Somaliland, and tensions were expected to rise.a In 2014, Puntland temporarily cut ties with the central government after the latter had signed an agreement with the Galguduud and Mudug regional authorities to form a federal member state despite Puntland’s (Box continued next page) Framing Subnational Governance to Secure Peace 47 CASE STUDY 3  Somalia—State Collapse and Loss of Monopoly on the Use of Force   (continued) claim for control over parts of the Mudug territory. Moreover, divisions over the formation of Jubaland exacerbated tensions among political elites at the national level and worsened relations between clans since 2013. At the regional level, political disputes emerged within the administration of Galmudug, which resulted in the signing of a power-sharing agreement to establish the offices of a state president, a state vice president, and a chief state minister, aimed at ensuring political inclusion for the various clans based in the Galmudug region. Another highly contentious, unresolved question in the state-building process is the status of the capital. Powerful interests in Mogadishu and its surrounding territory want to assign the capital a federal member state status and give it representation in the Upper House of Parliament, while others propose that Mogadishu should have a special status as a city-state or autonomous territory. As the central government still seeks to strengthen its authority over the national territory, large swaths of the country remain ungoverned by formal state authorities, and security forces are largely absent in rural areas. The central government has relied on external support and protection from forces of the United Nations–assisted African Union Mission in Somalia (AMISOM), which was established in 2007 and has been a challenging relationship to navigate. Somalia’s intrastate conflict was also exacerbated by the proliferation of militias that penetrated the country and retained control of large strips of the capital. Without a government that could provide security, communities relied on their clan affiliations for protection, which added to the militias’ raison d’etre and protracted the conflict. The collapse of the state and the inability of transitional governments to provide security and other basic services in the 1990s led to the rise of fundamentalist Islamic insurgent groups, most notably al-Shabaab. In addition to al-Shabaab’s advocacy of installing Shari’a law, the group’s ideological platform centers on its opposition to the central government and the West. While much of the territory previously under al-Shabaab’s control has been reclaimed by AMISOM and government forces, the group remains a serious threat to security in most parts of southern and central Somalia. Due to grievances felt by local communities, al-Shabaab is able to sustain recruitment and maintain a significant presence in southern Somalia. Historically, there have been tensions between consolidating political power through a strong centralized authority and employing the traditional pastoral kinship system that was characterized by the diffusion of power. Due to the legacy of the overcentralized Barre regime, many Somalis see federalism as a suitable model for a context in which state-society relations are defined by distrust of central authority. As far back as 1993, subnational governance was at the center of the peace negotiations held in Addis Ababa. However, the model was opposed by warlords, who viewed subnational governance arrangements as a threat to their political power. While the country is slowly overcoming these challenges with the current federal government, regional authorities continue to express fears of an overly “centralist” interpretation of the provisional constitution. Note: a. International Crisis Group 2018a. 48 Subnational Governance and Conflict Subnational Governance to Strengthen State-Society Relations Subnational governance can serve as a tool of state penetration and to strengthen state-society relations by providing public services in the periphery. In postconflict states, facilitating the functionality of institutions that can provide service delivery and quick peace dividends—as per an agreed-upon political s ­ ettlement—is paramount to sustaining stability and preventing renewed conflict. A recent study by the UNDP (2016, 3) concluded that “securing the social contract between state and society at the local level is essential to the long-term success of peacebuilding, statebuilding, and recovery processes.” Efforts to build the capacity of subnational governments must aim to strengthen and reform the social contract. Subnational governments rely on strong relations with the central governments to function effectively, which are in turn fundamentally defined by the nature of a state’s intergovernmental system. Subnational govern- ments have to coordinate closely with the central government in the implementation of both national and subnational policies. In turn, subnational governments must be allocated the necessary administrative authority and financial resources to imple- ment these policies and to meet the local population’s needs. This represents a partic- ular challenge in fragile and conflict-­ affected states, which face institutional weaknesses and often lack the requisite resources to strengthen the capacity of sub­ national governments. Compounding the problem is the fact that violent conflict and insecurity also trigger an exodus of civil servants and skilled labor, further impeding the ability of subnational governments to function (UNDP 2016). Ineffec- tive local governance is therefore at the heart of a vicious cycle in which violence and insecurity undermine the rule of law and institutional capacity, rendering it ineffec- tive for resolving disputes or curbing violence (World Bank 2011), and further exac- erbate violent conflict because of the failure of government institutions and the law to manage conflicts (World Bank 2017). Thus, particularly in fragile and conflict-­ affected settings, it is critical that subnational governments exhibit the capacity to resolve disputes peacefully. State-society relations are shaped by citizens’ perception of the legitimacy of the state, which in turn is informed by the state’s capacity to function effectively. Effective, resilient, and inclusive subnational governance is key to building a strong social con- tract between the state and society. State-society relations are particularly weak in conflict-affected or postconflict environments, and are typically characterized by dis- trust in and grievances against state authorities. Jackson (2016, 753) asserts that “the politics of centre-periphery relations are heightened in postconflict environments and include a general lack of trust, weak central authorities, financial pressures, political rivalry, and creeping centralisation of power.” In such settings, local communities pre- fer local governance that is predictable and sustainable—even if paired with negative externalities and undemocratic practices—over arbitrary and short-term engagements with democratic institutions (Felbab-Brown et al. 2018). When formal government institutions fail to extend state authority or provide basic services, non-state actors, including armed militias, step in and develop informal Framing Subnational Governance to Secure Peace 49 systems to fill the void. Some scholars note that in countries where the central govern- ment lacks the capacity to effectively establish state authority in the periphery, tradi- tional mechanisms of governance may be more effective at delivering services (Justino 2017a). In Mali, where the government is unable to provide basic services in the sparsely populated and vast territory of the North, non-state actors provide an infor- mal network for such services. However, non-state actors’ approaches to local governance is based on their own strategic interests, rather than on a desire to contribute to the public good. Felbab-Brown et al. (2018) found that the degree to which armed non-state actors provide governance and services to gain the local population’s support, versus the degree of coercion used to control the local population, is informed by these actors’ cost-benefit analysis of how long they can expect to maintain control. Armed non- state actors expect long-term returns for investing in local governance. Conversely, expectations of short-term control are linked to greater levels of brutality and violence. Even in cases in which armed non-state actors have held territory for a long time, the level of brutality and violence rises and service delivery decreases if forces expect to lose the territory within the foreseeable future. In turn, local communities’ support for non-state actors is determined by their perceptions of the time frame of the non-state actors’ withdrawal and the central government’s interest in and capability of retaking territory within a said period of time (Felbab-Brown et al. 2018). In postconflict settings, the provision of security, the rule of law, and basic services are major concerns for local constituencies. To maintain control over a territory, both state and non-state actors must enjoy a degree of legitimacy among the constituencies they seek to govern. Those state and non-state actors seeking to exert authority by strengthening state-society relations derive their legitimacy from the effective delivery of basic services—namely, the social contract. Some scholars suggest that when state and traditional actors compete for legitimacy at the local level, it results in improved service delivery and enhanced accountability (Mascagni 2016). However, Wall (2016, 910) cautions that “state legitimacy is built not just on which services the local councils deliver actively and inclusively, but also [on] the perceptions and expectations of these services.” Citizens’ perceptions of the strength of the social contract and the effectiveness of local governance are important predictors for the likelihood of intrastate conflict and communal violence (Justino 2017b). Predatory behavior by regional and local political leaders, along with a local community’s subse- quent frustrations over a lack of basic services—which are often attributed to senior officials’ unwillingness to serve their constituents—are persistent drivers of political unrest and violent conflict (Gaynor 2016). In fragile and conflict-affected settings that are defined by political patronage, improved local service delivery may benefit groups that are aligned with the political elite more than other groups (Justino 2017a). In Myanmar, for example, mismanage- ment of public institutions remains a major challenge for local governance, which cre- ates a negative perception of the state’s legitimacy and delivery of services. In addition, because some rural non-Burman people rely in part on patron-client relations with armed non-state groups to provide security and other basic services, conflict in Myan- mar is strongly shaped by local support for opposition groups. 50 Subnational Governance and Conflict Certain groups’ exclusion from accessing basic services and public goods is most apparent at the local levels (UNDP 2016). This has the potential to represent a destabi- lizing factor, as perceptions of injustice undermine social cohesion (Marc et al. 2012). It is at the local level that institutional capacity is the most relevant for addressing horizontal inequality and marginalization and for mitigating local drivers of conflict. After a political settlement with subnational governance provisions is successfully negotiated, local constituencies harbor high expectations that subnational governance reform will expand service delivery and that, as a result, their living conditions will improve. Similarly, subnational governance units have expectations of the central gov- ernment, in that they anticipate support in the form of financial resources and capac- ity building. Failure to manage these expectations can nurture grievances among all stakeholders and potentially reignite conflict (UNDP 2016). Some political settlements include provisions on the creation of special mechanisms that aim to provide quick peace dividends and demonstrate the positive impact of the settlement to local constit- uencies. In the Philippines, for example, a Special Zone for Peace and Development was created under the 1996 political settlement between the central government and the Moro National Liberation Front (MNLF) to provide economic guarantees and support basic service delivery and development so that a “peace dividend” would quickly materialize. Particularly in postconflict environments, subnational governments often lack the institutional and financial capacity needed to uphold the functions foreseen in politi- cal settlements and deliver much-needed services (Jackson 2016). Lack of capacity is further exacerbated by fragmented governance and a lack of coordination between different government entities—both within subnational entities, and between the cen- ter and the local levels. Considering that ongoing conflict and political instability deteriorate local constit- uents’ economic and social status and leave them without opportunities to secure their livelihoods, improving the delivery of public services is crucial in fragile and conflict- affected settings. Subnational governance is regarded as a viable tool for delivering public services more efficiently because it: (1) manages the delivery of services at dif- ferent government levels, depending on production at an optimal scale (whether large or small); (2) allows local governments to inform their policies based on the local con- stituency’s preferences and needs; and (3) creates competition between various gov- ernment entities (Lake and Rothchild 2005). It is naturally easier to formulate policies for a homogenous subset of society in a specific geographic territory than it is at the national level, where the entire state must be taken into account. Therefore, improved local governance as a mechanism to strengthen state-society relations is important for mitigating conflict and fostering stability. Moreover, in fragile and conflict-affected settings, social and economic deprivation can trigger communal violence over alternative sources of income and scarce resources. In this context, subnational governance can mitigate ongoing conflict and reduce the risk of renewed escalation of violence when used as a tool to deliver services, foster community cohesion, and build public trust (Wall 2016). Subnational governance is also expected to mitigate conflict by responding to the political, cultural, social, and economic needs of ethnically diverse societies. Framing Subnational Governance to Secure Peace 51 In contrast, Justino (2017a) found that community-driven local governance reform has limited effects on desired outcomes, such as social cohesion, collective action, and inclusiveness. Fragile, conflicted, and violent contexts are mired in distrust, and such interventions fail to take this fact into account (Justino 2017a). Justino (2017a) further contests that the values such initiatives seek to promote—power sharing, p ­ articipatory decision making, and ­ inclusiveness—cannot be achieved in fragile and conflict- affected settings, since disputes may arise over the very shape such arrangements may take. This shows that subnational governance arrangements must be based on local con- texts rather than imposed by the central government. In the context of long-established and manifested traditional governance systems, hybrid governance models can serve to bridge and consolidate formal and informal institutions for local service delivery. Marc et al. (2012, 188–89) identify “strengthening the interactions between formal and cus- tomary institutions as a means of improving the state-society relationship.” Particularly with regard to resolving local disputes, communities tend to demonstrate a strong pref- erence for customary institutions rather than formal ones. This preference is even stron- ger when there is a mismatch between formal and informal institutions, since the latter tends to better reflect a local community’s way of life (Marc et al. 2012). In Mali, subnational governance did not contribute to higher degrees of confidence in the state; instead, it destabilized and eroded the power of traditional authorities. Throughout Mali’s history, traditional authorities have played a critical role in manag- ing community life and access to resources and in resolving intra- and intertribal con- flicts. In the event of unresolved conflict, the traditional leader was normally the first to be informed, and his proposed solution was generally accepted by all parties. How- ever, the first republic abolished traditional authorities altogether. While these roles were reinstated in the 1970s, traditional authorities have struggled to redefine their formal role within the modern Malian state. Over the decades, the gradual deteriora- tion of this system in favor of formal government administration has largely created a vacuum in local conflict resolution and arbitration that the state has been unable to fill. Some governance interventions might only be suitable to the context of a particular region or local government and cannot be introduced across a state’s entire territory (Justino 2017a). In this context, subnational governance can serve to develop pockets of peace and stability by extending support to a few identified local-level government enti- ties. International development actors have successfully implemented this model in a number of fragile and conflict-affected settings. Strengthening subnational governance to underpin the social contract to in turn improve state-society relations across the entire territory, or in parts of it, is a prereq- uisite to fostering state legitimacy and setting the foundation for lasting peace and sustainable socioeconomic development. Conversely, a breakdown in state-society relations will herald calls for separatism, since certain segments of society do not per- ceive the state as a governance institution that represents their constituencies. 52 Subnational Governance and Conflict CASE STUDY 4  Myanmar—Renewing the Social Contract In Myanmar, subnational governance serves as a tool for state penetration in the periphery, aimed at building legitimacy through a renewed social contract. With minority groups calling for greater political, administrative, and fiscal autonomy, subnational governance institutions and state-society relations remain at the heart of the world’s longest running intrastate conflict. In recent years, Myanmar has embarked on a triple transition: from an authoritarian military system to a democratic governance system; from a closed to an open economic system; and from conflict to peace. Since around the time of Myanmar’s independence, ethnic minority groups have sought to address their grievances through demands for outright independence at first, and then later, increased local autonomy through federalism or other forms of subnational governance. Myanmar’s ongoing conflict has its roots in the 1948 struggle for independence and in a legacy of autonomy struggles for the borderlands, which were governed separately and largely autonomously as “frontier areas” under colonial rule. Minority groups called for a return to the spirit of the Panglong Agreement of 1947, according to which—and in return for joining the independent Burmese state—the Shan, Kachin, and Chin areas were granted full autonomy within the state and equal wealth sharing, while Shan and Kayah were offered the possibility of secession after a 10-year period. Following the coup d’état in 1962, the authoritarian governance style of the military junta left ethnic factions feeling forcefully integrated in national politics and unable to pursue self-governance. In 2003, under pressure from the international community, the military proclaimed a renewed effort to promote democracy by initiating a constitution-drafting process, but the process was not effective, with many groups boycotting it. A constitutional referendum was only announced in 2008. The 2008 Constitution provided for some decentralization of power but maintained a fairly centralist conception of the state. Accordingly, the subnational governance system comprises 15 administrative subdivisions, including one union territory, seven states, and seven regions. Regions are defined by the geographic concentration of the Burmese people, while states are dominated by minority groups. The regions and states are further divided into 74 districts, which are subdivided into townships, wards, and villages. These arrangements were elaborated in the 2012 Framework for Economic and Social Reforms, which called for a comprehensive decentralization policy, underpinned by new laws and regulations to support increased autonomy of states and regions. The Thein Sein administration, which came into office in 2010, made significant strides toward ending the conflict. In 2012, the country held its first by-elections with the main opposition party, the National League for Democracy (NLD), participating for the first time as an officially recognized party and winning a significant share of parliamentary seats. The NLD again dominated in the 2015 elections, and its leader, Aung San Suu Kyi, has subsequently been nominated as the de facto head of the government through the creation of the position of “state counselor,” which sought to bypass a constitutional clause preventing her from holding the presidency. (Box continued next page) Framing Subnational Governance to Secure Peace 53 CASE STUDY 4  Myanmar—Renewing the Social Contract  (continued) Recent ceasefire agreements aspired to end the more than 60 years of conflict. They involved negotiations with over a dozen different armed non-state actors (including the Karen National Union) that collectively control significant parts of territory, natural resources, and armed non-state actors throughout the country, and ended the world’s longest-running civil conflict. The Nationwide Ceasefire Agreement (NCA) was signed between the government and eight armed non- state actors in 2015, with two more groups signing in early 2018. The ceasefire discussions and the NCA itself have clearly set a federal vision for the state’s future structure, but what form federalism and wealth-sharing arrangements will take remains unaddressed. At least 10 ethnic armed groups have not yet signed the NCA. Despite pledges to strengthen the democratic federal union, the governance reform and political process have stalled under the NLD leadership. There has been little follow-up on the outcomes of the second Union Peace Conference held in May 2017. Consensus was reached that the state would be a federal democracy, but parties were unable to agree on divisive questions about self-determination and the right of some regions to adopt their own constitutions at the federal level. The third peace conference, hosted in July 2018, was viewed as exclusive and failed to live up to people’s expectations. The Rakhine crisis of August 2017, in which 725,000 self-identified Rohingyas were forced over the border to Bangladesh, was a further setback. There has been no progress on repatriation or on dealing with the root causes of the crisis, such as the lack of citizenship of many Muslims and restrictions on freedom of movement. From early 2019, violence increased in Rakhine with the expansion of a Buddhist ethnic armed group, the Arakan Army, further diminishing the prospects of refugee returns. Fighting also escalated in early 2018 in parts of northern Myanmar, and overlapping territorial claims by armed non-state actors continue to exacerbate localized conflict. The sheer number of armed non-state actors presents a unique challenge in terms of varying levels of state control over territory in the periphery. Much of the mountainous borderlands have never fully come under the authority of the central state; they were largely autonomous in precolonial times, were administered separately as semi-self-governing “frontier areas” during the colonial period, and fell under the control or influence of non-state armed actors in the decades since independence. The fact that many peripheral areas have never been under the central state’s control has reinforced the high level of distrust toward the state. There is a strong sense that the ceasefires have resulted in unwelcome government expansion into subnational areas that are subject to the settlements. For many, concerns are expressed over an unease with large-scale development, including existing local service delivery and administrative systems, because these represent a threat to the sovereignty of ethnic minority groups. For example, in some areas, new roads may also serve a military function by providing easier access for the Myanmar Army. To the extent that state authorities were present in these areas, they were seen more as resource extractors than service providers—or even as an occupying (Box continued next page) 54 Subnational Governance and Conflict CASE STUDY 4  Myanmar—Renewing the Social Contract  (continued) force, since they were mostly military rather than civilian. This means that the challenge of integrating the periphery into the formal central state system is not primarily a logistical challenge of extending state structures and service-delivery mechanisms into remote, hard-to-reach areas; rather, it is the much more difficult challenge of building confidence as part of an as yet inconclusive peace process. The current government has made significant concessions to transfer political and administrative power to the subnational level. As per the 2008 Constitution, steps have been taken to establish subnational legislatures, engage local authorities in budget allocations, and increase fiscal transfers to states and regions. The establishment of state- and regional-level legislative and executive bodies represents a significant move toward greater autonomy and accountability for local citizens. However, the military still holds one-quarter of the legislative seats at the state and regional level. The dependence of subnational governance units on central finances significantly restricts the exercise of democracy and subnational autonomy. The budget of the state and regional governments represents less than 12 percent of the entire budget, and public expenditure is largely controlled by the central government. Subnational governance is primarily realized through deconcentration, in which additional competences and budget allocations are transferred to the state and regional departments of central ministries and to the chief minister, but these actors remain largely accountable to the center, rather than to local communities. While the constitution stipulates that states and regions should enjoy exercising decentralized powers, their competences remain unclear, and relations between them and the central government are highly ambiguous. The General Administration Department has been transformed into a civilian ministry that is responsible for the appointment of chief ministers of the states and regions, which are in turn ultimately accountable to the president. This has fueled popular grievances. In this sense, public administration at the subnational level maintains the strongly centralized features of the pre-2008 period. The creation of new subnational institutions alongside older, centralized agencies has contributed to challenges in establishing management and accountability for services at the subnational level. Myanmar’s military also retains a veto on constitutional amendments. In 2019, acknowledging widespread concerns about the new constitution, the government established a review committee for the 2008 Constitution. The new president, Win Myint, who was appointed on March 30, 2018, pledged support for reforming the state structure on the basis of federalism. Steps to reform the subnational governance system are mostly visible at the village level where some elections have taken place, but the transformation to a more representative political system at the local level is being undertaken gradually. Subnational governance is characterized by the intersection of formal and informal institutions. One of the more tangible manifestations of this expansion was the creation of many sub-townships, almost exclusively in former conflict areas—an administrative level not provided for in the constitution. While the central government has enacted reforms to strengthen subnational governance and institutions of center-periphery relations in recent years, the (Box continued next page) Framing Subnational Governance to Secure Peace 55 CASE STUDY 4  Myanmar—Renewing the Social Contract  (continued) reform process is at an early stage and faces significant challenges. On the ground in large areas of the country’s contested territory, de facto administration takes place via various hybrid or non-state systems, with non-state actors playing a key role in governing and providing basic services. As the government fails to reach large parts of the country and public service provision is limited, non-state armed actors are left to fill the vacuum and perform a range of state functions. Meanwhile, decentralizing government services has political implications that may impact the peace process, and disagreements prevail over constitutional provisions. Collectively, non-state actors control significant territory, natural resources, and armed forces. Subnational governance arrangements may thus support state legitimacy and help address some of the principal drivers of internal conflict by achieving better and more responsive local governance, as well as increased transparency, equity, sustainability, and justice in sharing power and resources, particularly natural resource revenues. Nonetheless, the reform process and the government’s efforts to strengthen state-society relations will take place against the backdrop of years of conflict and deep-seated distrust toward the central state. The future shape and direction of relations between subnational units of government and the union level are thus core parts of Myanmar’s ongoing peace process, which aims to find an end to the multiple conflicts with ethnic armed groups that have affected the country since independence. Subnational Governance as a Slippery Slope to Secession? A failure to strengthen state-society relations, design inclusive political power- and wealth-sharing arrangements, or to extend state authority to the periphery often results in growing demands for independence. Since the end of the Second World War, more than 80 territorially concentrated ethnic groups have engaged in violent conflict demanding greater autonomy or independence, with varying degrees of success. For example, Somaliland announced its independence from Somalia in 1991, but has not been able to gain international recognition as a sovereign state. Meanwhile, Puntland has gained asymmetric regional autonomy, but in recent disputes with the central gov- ernment has threatened to follow suit and declare independence for the region of Puntland. Over time the feud has quieted, and Puntland remains a federal member state within the federal system of Somalia. Similarly, the Republika Srpska in Bosnia and Herzegovina has at times advocated for secession (and unification with neighbor- ing Serbia) but to date remains part of the state of Bosnia and Herzegovina. Identity groups often make secessionist demands by referencing the principle of self-determination enshrined in international law. However, international law fails to provide a clear definition and legal meaning of the term self-determination, or to pre- scribe pathways to attain it, leaving room for wide interpretation, including that 56 Subnational Governance and Conflict secession may lead to independent statehood (or unification with an existing state) or that decentralization may lead to greater self-governance within the boundaries of a state. This notion has conflicted with the question of how minorities’ right to self- determination infringes on the rights of other people—in such cases, the majority. The right to self-determination is also in direct contradiction to the legal norms of sover- eignty and territorial integrity (Mansell and Openshaw 2013), which central govern- ments cite in their opposition to secessionist claims. Secessionist conflicts have been identified as the most intractable form of intrastate conflict. Given the finite nature of a sovereign state’s territory, the central government is unlikely to cede parts of its national territory to a separatist group. Especially in cases where various identity-based groups are geographically distributed and expected to launch parallel secessionist movements, the central government will be reluctant to grant one separatist group’s demands for fear of encouraging other separatist groups. Instead, the central government will be inclined to signal that separatist conflicts are costly and would be to no avail. Although the immediate costs of a protracted intra- state conflict that involves secessionist groups are disproportionately high, the central government would be unwilling to take the risk to “negotiate itself out of existence” by accommodating the groups’ various demands (Walter 2003, 138). For example, the central government of Myanmar, which is comprised of both remnant elements of the authoritarian military and the newly elected civilian government, is confronted with a range of ethnic armed groups, most calling for self-determination to be realized in varying degrees of autonomy and in some instances even secession. Under the new constitution, Myanmar is constituted with a governance structure that is based on a unitary system with decentralization, but fails to grant any form of genuine self- governance or asymmetric autonomy. To a lesser extent, this also applies in cases like Bosnia and Herzegovina (demands from both Serbs and Croats) and Yemen (southern and Houthi separatism). However, while intuitively plausible, more recent research has cast doubt on such a domino effect, finding no evidence that concessions to one sepa- ratist group trigger an avalanche of similar demands by other groups (Forsberg 2013). Moreover, there is evidence that concessions by the central government regarding subnational governance arrangements can have a conflict-mitigating effect by making the attainment of a new political settlement more likely. In cases in which the central government is confronted with secessionist demands from only one distinct minority group, the cost of reaching a political settlement with some degree of asymmetrical subnational autonomy is lower than the cost of protracted conflict (Walter 2003). In the Philippines, for example, the MNLF gave up its demands for secession in return for greater autonomy, while the central government ceded political power to maintain its territorial unity. The political costs endured by both parties served as a signal for their commitment to sustained peace (Hartzell and Hoddie 2007). Comparative analyses also reveal that countries that expect a rise of secessionist movements preemptively grant a higher degree of fiscal decentralization, partly confirming that early concessions on subnational governance arrangements can have conflict-preventing effects (Cederman et al. 2015). Thus evidence suggests that states confronted with secessionist challenges respond strategically by devolving power through decentralization to avert the violent escalation of intrastate disputes over territory before they emerge (Tranchant 2016). Framing Subnational Governance to Secure Peace 57 Regional autonomy is also expected to reduce tensions by limiting the points of transactions between former adversaries. As the likelihood for secessionist tendencies increases exponentially in cases in which the existence of natural resources coincides with the geographic concentration of identity-based groups, such groups engage in secessionist movements to either increase or decrease the level of horizontal inequality. Where identity groups reside in territory that benefits from high economic growth or the presence of natural resources, these groups call for greater self-governance to either raise their share of the resulting revenues through favorable subnational governance arrangements or to take full control of them through secession. In other cases, such as Mali and the Philippines, disenfranchised identity groups that face regional inequality and disproportionately high levels of poverty base their claims for regional autonomy and secession on historical economic grievances (Ross et al. 2012). The literature reveals a lack of consensus among scholars on whether federalism and regional autonomy represent a step toward secession or whether the reverse is true—that federalism and regional autonomy serve as a measure to prevent secession- ist aspirations. Subnational governance arrangements are expected to avert secession by mitigating against entrenched polarization between and within minority and majority communities and against the entrenchment of grievances that can result in an emergence of radical political voices that drown out calls for dialogue by moderate figures. Consequently, regional autonomy with meaningful levels of political and fiscal decentralization is viewed as an effective tool for mitigating and ending secessionist conflicts. On the spectrum of subnational governance arrangements, asymmetric autonomy is most often granted as a measure to address secessionist demands, whereby specific territory is awarded greater autonomy than the remaining subnational gover- nance units (Weller and Nobbs 2010). Asymmetric autonomy for a specific territory is considered more effective in reducing and mitigating intrastate conflict when margin- alized or identity-based groups are geographically concentrated (McGarry and O’Leary 1993). However, the emergence of intrastate conflict as a consequence of secessionist aspirations is much more likely when minority identity groups are geographically con- centrated within a specific territory rather than dispersed across the national territory (Jenne et al. 2007; Toft 2003). While Brancati (2006) found that political and fiscal decentralization were effective mechanisms to dampen secessionist aspirations, she also cautioned that subnational governance arrangements that promote regional autonomy have been shown to encourage the formation of political parties that run on a platform of regional rather than national interest. Echoing these findings, Rode et al. (2018) concurred that terri- torial fiscal self-governance encourages regional parties to voice more radical demands in relation to the central government, since such political platforms gain overwhelm- ing support from the voter base in separatist conflicts. Jenne et al. (2007) found that historical experiences of groups having enjoyed territorial autonomy in the past are positively linked to separatist groups’ more extreme self-determination demands, but have no direct impact on the likelihood of violent conflict per se. Subnational gover- nance in the form of federalism appears to have a mitigating effect on intrastate 58 Subnational Governance and Conflict conflict. Bermeo (2002) asserted that no secessionist conflict within a federal democ- racy has led to the secession of the contested territory. When looking at the link between subnational governance and secession, there seems to be a case for differentiating between symmetric and asymmetric designs of fiscal decentralization. Rode et al. (2018) suggested that symmetric fiscal decentralization— granting fiscal autonomy to all regions—appears to have some effect on secessionist aspirations and should be considered a mechanism for mitigating secessionist conflicts. Meanwhile, by constituting a special status for a certain territory, asymmetric fiscal decentralization encourages secessionist demands, rather than dampening such senti- ments (Rode et al. 2018). An increase in fiscal transfers to a certain territory may enable the respective regional government to build sufficient governing capacity that would enable it to function outside the realm of the nation-state. This is another indication of the concern regarding the risk that subnational gover- nance arrangements may prepare subnational governance entities for secession and thereby increase the potential for secessionist movements. Channeling state resources to autonomous regions enables regional administrations to build the institutional capacity necessary to govern a sovereign state. As autonomous regions gain more capacity for self-governance, separatist groups use their special status as a negotiating tool with the central government (Chapman and Roeder 2007; Rode et al. 2018). In a more radical approach, arguments have been made in favor of secession as a measure to mitigate intrastate conflict fought along identity lines. In this logic espe- cially, ethnic conflicts could be resolved by physically separating opposing groups, thereby reducing the risk of escalating violence. Somaliland appears to have benefited from effective secession, which came about in the wake of a civil war that centered on a power struggle over control of the central government. Proponents of this approach reference the security dilemma faced by minority groups in situations of political instability. This concept proposes that minority groups and marginalized communities are expected to initiate a preemptive war as a defense strategy against surrounding groups (Posen 1993; Lake and Rothchild 1996). In this context, secession is expected to mitigate the security dilemma by separating heteroge- neous groups and creating clear boundaries of defense (Sambanis and Schulhofer-Wohl 2009). From this perspective, secession of contested territory is viewed as an effective measure to prevent a resurgence of violent conflict, since each of the former adversar- ies now holds its own territory and is protected by internationally recognized borders (Kaufmann 1996). However, Sambanis (2000) found that secession is a very costly endeavor, which in turn encourages warring parties to develop institutional capacity to resolve disputes peacefully within existing state borders. Furthermore, once a region has seceded, there is a serious risk that new identity-based groups will form within the newly indepen- dent territory. As Sisk (1996, 2) pointed out that in such contexts, subnational gover- nance arrangements merely “rearrange the configuration of minorities and majorities.” Accordingly, accommodating minority groups through regional autonomy or seces- sion may ensure that newly formed minority groups are marginalized (Sisk 1996). These so-called orphans of secession (McGarry 1998) are expected to engage in a renewed struggle for political power and access to state resources, risking the Framing Subnational Governance to Secure Peace 59 escalation of violent intrastate conflict within the territory of the new sovereign state and potentially another round of state fragmentation. An example of this comes from the post-Dayton demands by the Republika Srpska to be allowed to secede from Bos- nia and Herzegovina. Thus, in general, there is broad consensus in the literature that secession is not a suitable tool for managing intrastate conflicts that have a subnational dimension (Jenne 2012; O’Leary 2007; Sambanis 2000; Sambanis and Schulhofer-Wohl 2009). Therefore, Sambanis and Schulhofer-Wohl (2009) concluded that secession is not likely to reduce the chance that intrastate conflict will recur, and might actually increase it. This demonstrates the need to strike a balance between granting a degree of auton- omy that accommodates minorities while also maintaining the state’s territorial integ- rity. Some political settlements that grant asymmetric regional autonomy include explicit language that prohibits the autonomous entity from promoting secession. CASE STUDY 5  Bosnia and Herzegovina—Administrative Walls of Distrust The subnational governance arrangements adopted by the 1995 Dayton Peace Accords aimed primarily to end the war and ensure the survival of Bosnia and Herzegovina as a state, which in turn had to accommodate the strategic interests and requirements of all warring parties and their external patrons. A highly inefficient governance system and a very complex political system were accepted as the costs of maintaining peace and the country’s territorial integrity. After the disintegration of the Socialist Federal Republic of Yugoslavia, Bosnia and Herzegovina became engulfed in a civil war from 1992 to 1995. Elections held in 1990 resulted in a parliament sharply divided among the three leading ethnic groups—Bosniaks, Serbs, and Croats—especially over the question of independence. When Bosnia and Herzegovina declared independence in 1992, the Serb population refused to accept the outcome of the referendum and announced aspirations to create a separate state in the territories with ethnic Serb majorities. Following the referendum, local Serb militias, supported by elements of the Yugoslav People’s Army, escalated violence that rapidly spread to involve all three ethnic groups and transformed into a full-blown war. Seeking to secure Serb majority territory, Serb forces fought the armed forces of the Republic of Bosnia and Herzegovina, which was initially composed of all three of the main ethnic groups. However, secessionist Croats raised territorial claims, but were later integrated into the Federation of Bosnia and Herzegovina. As a result, war ignited on another front between Bosniak and Croat forces. Thus, in postwar Bosnia and Herzegovina, the Bosniaks, who are mostly Muslim, represented a plurality of about 50 percent, while the Serbs, who are mostly Orthodox Christian, represented roughly 30 percent, and the Croats, who are mostly Catholic, represented about 15 percent. In this ethnically polarized context, subnational governance provided a framework for interethnic accommodation. Both the Serbs and the Croats advocated for a highly decentralized system in which the governing authority lies within territories where each group held a majority and expected to retain substantial political (Box continued next page) 60 Subnational Governance and Conflict CASE STUDY 5  Bosnia and Herzegovina—Administrative Walls of Distrust  (continued) support. Meanwhile, the Bosniaks sought to assign the central government significant competences, as they expected to hold a majority at the national level.a As outlined in Annex 4 of the Dayton Peace Accord, which outlined the new constitution, Bosnia and Herzegovina was divided into two entities—the Republika Srpska and the Federation of Bosnia and Herzegovina—and further layers of subnational governance units. Brcko District was added as another self-governing administrative unit in 1999. The de facto separate governments and administrative organs enjoy substantial legislative and executive powers. For the two entities, the constitution has set up an asymmetrical system between them, since each entity demanded a different degree of subsidiarity. The Federation of Bosnia and Herzegovina contains 10 intermediate-tier units (cantons) and 80 municipalities. Its constitution defines the functions of each level of government, including granting to the cantons all powers not expressly granted to the entity level. The cantons have their own legislatures, their own basic laws and constitutions, and their own governors and ministries. The Republika Srpska has a centralized administrative structure and its own ministry of local government to regulate and conduct dialogue with its 63 local governments. In addition, the Dayton Peace Accords proposed a number of other power- sharing and power-dividing mechanisms that are institutionalized at the central level but reflect a high degree of subsidiarity—as was required to get the parties to agree to remain part of the same state. These include: (1) an executive branch whose highest powers reside in a three-person rotating presidency based on ethnicity—one Serb to be directly elected from the Republika Srpska entity, and two (one Bosniak and one Croat) from the Federation of Bosnia and Herzegovina entity—and the Council of Ministers, which represents the central government; (2) a judicial branch whose highest powers reside in a constitutional court that rules on constitutional disputes, including those on appeal from entity courts; and (3) a bicameral parliament. Yet these central- or state-level institutions have only limited exclusive responsibilities, namely for defense and security, foreign policy, human rights and refugees, foreign trade, customs, and monetary policy. All government functions not explicitly assigned by the constitution to the state are assigned to the two entities. Given the sensitivities left from the war, the parties were loath to establish any system that involved fiscal interdependency. Identity-based politics and mobilization played a central role in almost every aspect of statehood, including election procedures. Overall, the system’s combination of subsidiarity and veto points has significantly hampered the ability to reach effective decisions. In the postconflict environment, minority groups’ fears of being dominated by a larger group, along with the tolerance for inequality in service delivery, were greater than that for cross-subsidies; there was even resistance to cross-subsidies between groups within the same multiethnic municipality. In some cases, where territorial-based fiscal federalism was not possible because of the intermixing of (Box continued next page) Framing Subnational Governance to Secure Peace 61 CASE STUDY 5  Bosnia and Herzegovina—Administrative Walls of Distrust  (continued) ethnic groups, the service-delivery system was established to benefit only certain ethnic groups, rather than the entire constituency living inside certain territorial boundaries. Both the better-off and worse-off areas preferred control of their own affairs against the possibility of having more revenue but less direct control. Each group believed it would lose out in any redistribution, and none trusted that revenues would be distributed in a transparent or equitable manner. The economic and political elements of subnational governance arrangements outlined in the Dayton Peace Accords were negotiated in parallel and separately, leading to some degree of incoherence. The accords devised a political rather than economic plan for devolving government. Therefore, the settlement did not represent the outcome of one integrated and holistic negotiation process. While the economic issues were in a sense the “least possible bad” options to which the three parties would agree, the political aspects were more the “best possible practice” that had little support from two of the three negotiating parties. In this case, the extensive degree of subnational governance sought to reduce the number of interactions in which disagreements between ethnic groups could emerge, rather than striving for an efficient governance model. Those in each group perceived themselves as members of their ethnic group rather than as Bosnians, and therefore conflict management sought to bring institutions of power closer to the people to ensure that decision making could be more sensitive to the needs of the different ethnic groups. Although the formula of extreme subsidiarity has resulted in a highly inefficient governance structure, it persuaded the different ethnic groups to agree to cohabitate within the same state. Bosnia and Herzegovina’s subnational governance system is essential for the stability of the country’s territorial integrity. However, the sustainability of the model itself is in question, since the continued existence of parallel structures and the emphasis put on avoiding interdependencies suggest that the Federation entity may also be at risk. The Croat minority there favor relatively strong cantonal governments, while the Bosniaks favor stronger entity-level institutions, which they may be able to control. The limitations of the subnational governance arrangements and the complex political system set up at Dayton came to light at the most recent presidential and parliamentary elections in October 2018. Bosnia and Herzegovina cast ballots for a broad range of representatives, including five presidents and members for the 14 legislative bodies, with each of the 10 cantons having its own parliamentary assembly. The Serb nationalist Milorad Dodik, who is calling for secession of the Republika Srpska and publicly announced that he would only advance the interests of the Serb community, emerged victorious in the election for the Serb seat of the three-member state presidency. Meanwhile, constituencies of the Federation of Bosnia and Herzegovina elected the Bosniak and the Croat members of the presidency. The moderate Zeljko Komsic won the Croat seat, prompting the Croat community to protest the election results, accuse the Bosniaks of out-voting the nationalist Croat candidate, and call for the formation of a separate entity with its own electoral constituency. In the elections held in 2000, Serb nationalists won the (Box continued next page) 62 Subnational Governance and Conflict CASE STUDY 5  Bosnia and Herzegovina—Administrative Walls of Distrust  (continued) largest share in the Republika Srpska entity, while moderate parties dominated in the Federation entity, which reversed in 2002 when nationalists regained power in the Federation entity’s presidential, parliamentary, and local elections. The 2010 general elections were followed by a more than one-year political deadlock that was resolved by an agreement reached between Bosniak, Serb, and Croat political leaders on the formation of a coalition government. The highly decentralized nature of the state agreed at Dayton made peace possible, but subnational governance as a method of effective and efficient governance is far from being realized. In February 2014, weak governance, political corruption, and mass unemployment triggered demonstrations and violent riots in the Federation entity. Nationalist elites have instrumentalized subnational governance as a means of ethno-centralization rather than as an effective tool for peace building and (democratic) state building. While the subnational governance arrangements designed at Dayton created a governance structure that was riddled with inefficiencies and blockages and resulted in a bloated public sector defined by ethnic division, the agreement nonetheless brought an end to years of bloodshed and has thus far ushered in a period of more than two decades of peace. Note: a. O’Brien 2010. What’s in the Law—Political Settlement Provisions The law manifested in constitutions or other forms of political settlements defines the state system and organizes and establishes political power. It does so by constituting and distributing state authority and political power among government institutions and between the state and citizens, by assigning competences for policy making and awarding or constraining formal power (World Bank 2017). Therefore, a state’s consti- tutional framework and settlement provisions govern relations between central and subnational governance tiers and underpin proposed subnational governance arrange- ments with a legal framework. This study has adopted a definition of subnational gov- ernance that emphasizes the legal entrenchment of the exercise of subnational governance powers, as well as the overall prevalence of a state’s legal order. Therefore, provisions in the law simultaneously enable and constrain subnational governance, establishing guarantees both for self-governance at the subnational level and to pre- serve a state’s territorial integrity. For both of these guarantees to be meaningful, com- mitments to the rule of law—and effective institutions upholding it—are essential. Such commitments and institutions need to penetrate the entire governance sys- tem. Often at the subnational level, self-governing entities have an internal constitu- tion or a constitutional legal framework determining their own governance structure. In Bosnia and Herzegovina, the two subnational entities—the Federation of Bosnia and Herzegovina, and the Republika Srpska—are governed by two distinct subna- tional constitutions. Similar to the state level, ethnic divisions play an important role Framing Subnational Governance to Secure Peace 63 in determining the provisions of subnational constitutions, since these stipulate which ethnic groups are the “constituent” people of the subnational entity. In 2000, such pro- visions were ruled unconstitutional by the country’s constitutional court because they constrained the citizenship rights of those from a different ethnic background than the “titular nations” of each entity. However, since the Dayton Peace Accords also oblige Bosnia and Herzegovina to conform with certain international human rights instru- ments (such as the European Convention on Human Rights), a case brought before the Grand Chamber of the European Court of Human Rights found Bosnia and Herze- govina’s constitution itself to be in violation of the country’s obligations under inter- national law and demanded constitutional reform (Milanovic 2010). This reform process, however, is yet to be concluded. Typically in postconflict countries, subnational governance arrangements are man- ifested as legal frameworks in peace agreements, constitutions, or other parts of the legal system governing the state structure. Therefore, any reform process that changes a state’s subnational government arrangements must be preempted by redrafting the legal framework and introducing amendments to relevant legislation. In most cases, transitions from conflict to peace culminate in a constitution making process. In Bos- nia and Herzegovina, the constitution was outlined in Annex 4 of the 1995 Dayton Peace Accords. Legal Provisions That Reflect Power Balances and Local Realities For subnational governance to serve as a useful tool for conflict mitigation, the subna- tional governance arrangements and specific design of the political settlement chosen must address the grievances and interests of warring parties and relevant stakeholders and tackle the drivers of conflict. In cases in which a political settlement’s subnational governance arrangements leave certain groups with new grievances or fail to address deep-rooted ones, it is likely that intrastate conflict will recur. Political settlements are mostly negotiated by national-level political elites, giving subnational government officials or traditional authorities who enjoy legitimacy among local communities little say in the design of the arrangements that will funda- mentally alter their governance system (Jackson 2016). Successful subnational gover- nance arrangements have to take into account the grievances of both central- and local-level actors to effectively mitigate conflict. Assessments of country experiences suggest that in some cases subnational governance arrangements are designed to address national-level conflict dynamics, thereby failing to mitigate drivers of localized conflict (Gaynor 2016). In this vein, legal provisions that underpin subnational gover- nance have to reflect the complexities of subnational governance reform, taking into account not only national-, subnational-, and local-level conflict dynamics but also relations between the various layers of governance, as well as formal and informal gov- ernance mechanisms. The constitutional framework and legal settlement provisions also have to provide clarity on the relations between formal and informal governance mechanisms (Narang Suri 2016). Justino (2017a, 14) adds that “local initiatives which 64 Subnational Governance and Conflict take into account the myriad interests and power structures that emerge at the subna- tional level and their informal nature may yield positive outcomes.” Constitutional Guarantees for Minorities and Marginalized Groups Building on the recognition that third parties have a significant role to play in address- ing commitment problems, effective constitutional structures are key in providing leg- islative constraints to political power holders. Minority and marginalized groups that expect to lose political influence in the absence of an externally led political settlement process may advocate for extensive legal guarantees to protect their rights and political participation. Entrenching subnational governance in constitutional law provides guarantees for minority groups on the level of self-government they enjoy. Also, to prevent local elites from capturing local government, subnational governance arrange- ments must be underpinned with a clear legal framework that outlines the compe- tences of local leaders, and constituencies need to be able to hold their local representatives accountable. Unless they are based on reciprocal commitments by minorities and marginalized groups, there is a risk that guarantees and protections for minorities and marginalized groups will raise tensions in the long term, since majority groups may perceive them as a threat to the unity and survival of their state. Balancing Agreements between Ambiguity and Precision Fragile and conflict-affected settings require a sensitive balancing act between ambigu- ity (which is necessary to enable specific settlement provisions to be adapted to a changing context) and precision (so that powerful actors cannot interpret settlement provisions as they wish). Particularly in the context of ongoing conflict, international actors tend to rush warring parties to reach an agreement that lacks precision on con- tentious issues. Efforts to build consensus over divisive matters and to reach specific agreements may stall the negotiation process, lead to an escalation of violence, and potentially prolong conflict. Seeking to advance the process of engineering a new political settlement, parties may choose to introduce ambiguity to achieve an agree- ment that emphasizes core principles that all parties can rally behind, but is vague enough on operational details so that each party can claim it has achieved its own objectives. However, avoiding clarity on the most contentious matters and failing to include specific provisions risks ending up with a stalemate and renewed disputes in the implementation stage. In other words, such an approach merely postpones ten- sions and conflict to the post-agreement stage. For example, without enforceable clar- ity over the territorial delimitation of subnational governance units, postconflict state-building processes have been accompanied by heightened levels of violence as various actors fight for control over territory and contest internal boundaries (Mogaka 2017). In Somalia, the state-building process envisaged by the 2012 constitution has been mired in disputes between various regions over border demarcations of the newly established federal member states. The 2012 Constitution proposes that a minimum of Framing Subnational Governance to Secure Peace 65 two regions have to merge to form a federal member state, but failed to provide the necessary details to avoid the level of instability that initially accompanied Somalia’s state-building process. The 2012 Constitution also fails to provide guidelines on how to distribute competences between the central government and the federal states, out- lining only that the allocation of powers and resources has to be based on agreement between the central government and the newly formed federal member states. There needs to be clarity on the distribution of competences between central and subnational governance units, and on which entity holds policy making authority. Since constitutions or other political settlements that outline subnational governance systems are highly complex, it would not be feasible to address the separation of com- petences for all contingencies. Subnational governance systems involve a range of gov- ernment entities from the center to the local level, each holding their very own, and often conflicting, interpretation of legal provisions. This is particularly important in the context of intergovernmental fiscal arrangements, such as formulas that aim to equalize the distribution of state resources proportionally to population size or to address historical social and economic marginalization. To avoid future contestation over the interpretation of a new political settlement’s legal provisions, commitments to peaceful dispute settlement are as essential as the institutional mechanisms associ- ated with them and should form part of the settlement (Wolff 2013a). Here, too, inter- national development partners have an important role to play in helping conflict parties find the right formula and support the implementation and operation of the institutions and mechanisms associated with it. For example, under the provisions of the Dayton Peace Accords, a Human Rights Chamber was established within the Constitutional Court of Bosnia and Herzegovina, with six national and eight international members. The latter were selected by the Committee of Ministers of the Council of Europe. Implementation of decisions by the Human Rights Chamber were monitored by the High Representative and the Organi- zation for Security and Co-operation in Europe. The Human Rights Chamber ceased to operate in 2003, and was replaced by the Human Rights Commission, which also ceased to operate in 2004. Today, the Constitutional Court is responsible for all cases alleging human rights violations. The Constitutional Court currently consists of six national (two from each ethnic group) and six international members. A failure to include specifics may also lead to questions regarding the legitimacy of institutions or the legality of state actions, which in turn compromises both the politi- cal settlement and the subnational governance reform process. Nonetheless, in fragile and conflict-affected countries, where newly formed subnational governance entities often lack the institutional capacity to fulfil their mandates or where fighting contin- ues in certain areas, the assignment of competences must follow a staged approach, in which the allocation of competences for each level of government takes into account the local situation (World Bank 2014). Extensive ambiguity and a lack of granularity allow the various parties to interpret their commitments under the proposed subnational governance arrangements differ- ently and in a way more favorable to their constituents. Furthermore, vague subna- tional governance provisions enable parties to renege on their commitments; this is particularly the case if there are significant power imbalances between the warring 66 Subnational Governance and Conflict parties. There is a clear recognition that a legal framework is only as credible as the commitments underpinning it, which creates a shared understanding about the expected behavior of formerly warring parties and encourages cooperation to pro- mote positive outcomes (World Bank 2017). From Paper to Peace Legal provisions are based on the notion of norms. However, if the gap between the normative framework and the reality of the local context is too wide, the settlement might face implementation and credibility challenges. In Mali, the warring parties have signed a number of peace agreements that legally gave the North greater auton- omy. However, given that the national government has had significant challenges in devolving political power and that the financial resources underpinning the legal framework were insufficient to successfully implement the peace agreements, the nor- mative provisions failed to change reality on the ground. This case clearly demonstrates that agreements reached on paper have to translate into changing realities on the ground. Particularly in the case of proposed subnational governance reform, local constituencies hold high expectations about improvements in local service delivery and increased access to political decision making, both at the center and at subnational governance levels. So, while the legal framework may be clear on the nature of subnational governance arrangements, efforts to support the imple- mentation of such arrangements need to address potential commitment problems. CASE STUDY 6  Yemen—A Flawed Constitutional Framework Disagreements over the constitutional design of Yemen’s subnational governance arrangements were a key impediment for the transition from authoritarian rule to democracy, and the country eventually plunged into a devastating internationalized intrastate conflict in 2015. Some constituencies advocated for a strongly devolved federal system and some in the South even promoted secession, while others supported a highly centralized governance system. Yet all of the weaknesses in the original agreement (and those that followed) were less important than the deep divisions between the political elites, who failed to reach a settlement that would secure their commitment by accommodating everyone’s core interests. In 2015, Yemen experienced a return to civil war following a period of relative calm and considerable progress in state and peace building, which led to the emergence of two parallel governments: one in Aden and one in Sanaa, the capital. Inspired by Arab Spring events, mass demonstrations led to the resignation of then President Ali Abdullah Saleh and the collapse of his regime, which had been in power for over three decades. The sitting vice president, Abdrabbuh Mansur Hadi, was nominated to serve as president and launched the National Dialogue Conference (NDC). Yet in late 2014, the transition process began to unravel before a draft constitution that was based on the outcomes of the national dialogue could be implemented. (Box continued next page) Framing Subnational Governance to Secure Peace 67 CASE STUDY 6  Yemen—A Flawed Constitutional Framework  (continued) Under the 1991 Constitution, following the unification of North and South Yemen, the country was organized by a unitary governance system, with a large number of subnational governance units comprising governorates, districts, and subdistricts that merely enjoyed administrative competences. Subnational government officials were accountable to the central government and not to their local constituencies. In response to growing public discontent, the government developed a national decentralization strategy in 2008, but its implementation was limited. Local authorities received only about 16 percent of public expenditures, almost all of which were controlled by the governor, who was an official of the central government. Moreover, the state’s presence in much of the country’s periphery remained extremely limited. The formal structures of government were secondary to the almost feudal control that tribal sheiks exercised over their constituencies. There were two main constituencies actively contesting developments of the political transition. First, there was the southern separatist movement known as Hirak. The movement rallied around the deep grievances in the South over its treatment since unification in 1990 and, more particularly, since the civil war in 1994, which was defined by the marginalization of southerners in centralized decision making. The South, which represented 20 percent of the population, produced the majority of the country’s oil but felt it received little in return. Thus, Hirak was a territorially based political movement whose agenda was focused on demands for new arrangements with greater autonomy (or even independence) and more influence at the center. Second, the Houthis, based in the northern governorate of Saada and part of the Zaydi tribal group, had grievances about neglect by the regime but also about Saudi-sponsored Wahhabi proselytization in their area. Like Hirak, they demanded greater autonomy for their homelands. The newly formed National Reconciliation Government left the old party elites in power and completely excluded those who had pressed hardest for serious political change. Importantly, the new government excluded the Houthis, Hirak, and civil society groups associated with the Arab Spring demonstrations. The NDC was tasked with formulating the mandate for a constitution drafting assembly, to prepare a constitutional proposal that would be put to a referendum and serve as a legislative framework for elections. However, the timetable proved unrealistic because facilitating the NDC and drafting a constitution took far longer than envisaged. The public became progressively disillusioned with the stalling transition process. Consensus emerged within the NDC in favor of federalism, proposing a four- tier federal system with the center, governorates, districts, and subdistricts. But support for federalism was weak and failed to resolve any of the fundamental design questions on the allocation of powers and central institutions. It was unlikely that a conference of 565 people of widely different affiliations would reach consensus on a governance system or other contentious matters. At the time the event concluded in September 2013, it produced 1,800 proposals, including that Yemen should become a federation with six regions with Sanaa as an independent capital city. However, many of these proposals expressed aspirations with abundant contradictions, and few tackled divisive questions of federalism, the (Box continued next page) 68 Subnational Governance and Conflict CASE STUDY 6  Yemen—A Flawed Constitutional Framework  (continued) South, and the military. Success in such a deeply divided society, with little state capacity, deep poverty, and severe economic and security problems, was always going to be a challenge, but the design and execution of the transitional process, in which international actors played an important role, contributed to the failure. On the basis of these proposals, a constitutional framework was designed that provided for a federal system comprising six regions, dividing the South into two regions and the North into four regions. Hirak was infuriated, insisting on a united southern region. However, its more serious political consequence was the total rejection by the Houthis. Both factions favored a federation that would be based on the country’s division into two regions, rather than the six regions proposed in the constitutional draft. The territory encompassing the Houthi constituency was to be situated within the new region of Azal, with no access to the sea, little water— which is a scarce resource in Yemen and a significant driver of localized conflict— and few other resources. The Houthis would have been a minority in the region and wanted instead to be united with the neighboring Al Jawf governorate, which had better access to resources. They thus viewed the proposed design as an attempt to disempower and impoverish their constituency. Despite the opposition to this model by southern representatives in the NDC and by the Houthi faction, the president established an unrepresentative Constitutional Drafting Committee (CDC) in 2014, which presented a draft constitution in January 2015—an event that was viewed as a trigger for the final Houthi takeover of Sanaa and the transition to all-out civil war. The NDC process revealed the contentious nature of federalism, and many stakeholders were concerned about the CDC’s ability to establish a suitable federal arrangement. In addition to the six regions and two autonomous cities of Sanaa and Aden, the draft maintained the existing governorates (mohafazah) as a constitutionalized tier of government and provided for local districts, which would have required financial resources beyond Yemen’s capacity. The federation’s structural problems would have been further aggravated by the unclear division of competences, since some exclusive competences were assigned to both the federal and regional levels. The Houthi takeover, coupled with the fact that the CDC faced major challenges in translating the NDC proposal on the division of power into a constitutional framework, further strengthened calls for secession among some southern constituencies. The South in particular had grievances about how its region’s oil reserves were being managed. Southern oil-producing regions expressed discontent with the wealth-sharing arrangements, arguing that the central government and the corrupt political elite disproportionately benefited from natural resource revenues. An estimated 60 percent of oil production is concentrated in the South—as compared to 40 percent in the North (in Mareb region) —and the two oil-producing regions (Shabwah and Hadhramout) represent roughly 7 percent of the population. Solving the dispute over resource management received special attention from the NDC during the constitution drafting process. While the draft constitution called for a federal law on revenue sharing, fundamental issues remained unresolved. (Box continued next page) Framing Subnational Governance to Secure Peace 69 CASE STUDY 6  Yemen—A Flawed Constitutional Framework  (continued) Yemen has been highly centralized since unification in 1990, but the state had only a limited presence in many regions. In addition, President Saleh’s style of patronage politics progressively lost its ability to “dance on the heads of snakes” and could no longer maintain a balance between competing forces. The deeply alienated South, crushed in the 1994 civil war, sought empowerment but favored secession over federalism. It was the challenge from the South that led the NDC to consensus on the need for federalism, which President Hadi endorsed as a way to dampen southern secessionist aspirations. However, the Hirak faction initially refused to participate in the NDC process and never bought into Hadi’s six- region federal model. The Houthis, for their part, were not in principle opposed to federalism, but they too rejected the proposed subnational arrangements due to the regional design that failed to account for the group’s geographic location. The two leading established parties in the North were both centralist in orientation and thus resistant to the degree of devolution provided by a federal system that may have gained some support in the South. The continued exercise of parallel governance by the Hadi administration in exile and the Houthi-controlled government in the capital risks reinforcing regional divisions. While the Houthi government is not internationally recognized, its forces effectively control most of the Northwest, and the Parliament convened to issue a vote of confidence in 2016, whereas parliamentarians loyal to former President Saleh were able to reach a quorum despite the boycott of southern parliamentarians. Meanwhile, governors who were supposed to be elected by local councils have been appointed by both entities and lack legal legitimacy. In October 2018, in light of intensifying conflict accelerated by external military support from Iran for the Houthi faction and from a Saudi-led coalition for the Hadi government, senior United States and United Kingdom officials called on conflict parties to reach a ceasefire agreement. The United States, the United Arab Emirates, and other regional states are providing military assistance to the Saudi- led Arab coalition. Meanwhile, peace talks under the auspices of the UN have not yielded any results, particularly since the Houthis refused to leave the country to participate in the most recent peace talks in Geneva in September 2018. The exiled Hadi government has rejected any peace agreement that does not guarantee the full reinstatement of its government, while the Houthi faction insists it must have a seat at the table. While questions regarding the federal system are not at the center of ongoing peace talks, these issues are expected to reemerge as the key element in the transition and the constitution-making process. In December 2018, the UN was able to relaunch the peace dialogue in Sweden between the Yemeni government and the Houthis. The talks, the first in two years, resulted in agreements on a prisoner swap and a humanitarian truce in the cities of Hodeida and Taiz as preludes to a political framework for a lasting ceasefire and political transition going forward. Thus, subnational governance was at the heart of the political agenda, but the NDC consensus in favor of federalism was very fragile and in no way provided a basis for a real and lasting political agreement. As much as politics seemed to require federalism (complemented by power sharing at the center), the constitution making process was inadequate for the challenges of designing such complex arrangements. A failure to build consensus on the federal state structure among the various parties thus precipitated the civil war. 70 Subnational Governance and Conflict Addressing Commitment Problems After a political settlement is successfully negotiated, its provisions must be translated into tangible reforms and functioning institutions. Changing a state’s institutional structure is not merely a technical but an inherently political process. Acemoglu and Robinson (2006) add that reforming governance structures shifts the power balance between actors, creating winners and losers. This is fundamentally the case when reforming and strengthening subnational governance arrangements, since they reflect and institutionally reshape power balances between different political actors, both at the center and at subnational levels. Subnational governance reform is therefore a political decision, and the design of such arrangements is the outcome of a bargaining process between various parties that reflects their relative balance of power and com- peting political incentives (World Bank 2017). Consider the ongoing conflict in Myanmar, which began with the broken promise of the 1947 Panglong Agreement that granted the Kachin, Shan, and Chin ethnic groups full autonomy but was never formally incorporated in the 1947 Constitution. Similarly, the 1974 Constitution and the constitutional convention of 1993 failed to provide for subnational governance arrangements that would have allowed for mean- ingful democratic participation and local self-governance. This occurred despite the fact that poor governance and the military regime’s refusal to discuss political issues have been linked to the collapse of a number of ceasefires reached in the late 1980s and early 1990s. Subnational governance arrangements as conflict mitigation mechanisms seek to address concerns by marginalized and excluded groups by rebalancing power and resources. A majority group’s declaration to consent to subnational governance arrangements in the form of power sharing and dividing sends a costly signal to its adversaries that it is willing to cede power in the interest of achieving peace and stabil- ity (Roeder and Rothchild 2005). Subnational governance arrangements that allow marginalized groups to participate in decision making and hold local representatives accountable reduce the opportunity for central state authorities and political elites to renege on their commitments. However, the available evidence is not conclusive on the extent to which subna- tional governance arrangements positively contribute to institutional structures that reflect the prevailing balance of power between relevant groups and sufficiently accommodate their respective interests, therefore offering reasonable prospects for political stability. Yet neither is there conclusive evidence that subnational governance arrangements merely offer a short-term, temporary opportunity to terminate violence while increasing the likelihood of conflict recurrence. This is a perennial debate between advocates and opponents of subnational governance arrangements. The for- mer argue that such arrangements accommodate and integrate challenger groups through a new political settlement that acknowledges and addresses their grievances. In contrast, the latter argue that subnational governance arrangements simply provide additional political and economic resources, which empowers such groups to chal- lenge this new political settlement again in the future. Framing Subnational Governance to Secure Peace 71 In most cases, intrastate conflict is defined by significant power imbalances between the central government, which has access to resources and the national security appa- ratus, and non-state armed actors in the periphery. However, power relations between central and peripheral government entities, as well as between state and non-state actors, shift substantially during wartime (Jackson 2016; World Bank 2017). As negotiations on the nature of the state and the degree of subnational governance occur within the framework of elite bargains, attempts to analyze the potential of sub- national governance reform on intrastate conflicts with subnational dimensions has to take into account the unequal power balance between central and peripheral actors. When seeking to address commitment problems in the implementation of subnational governance arrangements, stakeholders must question the incentives of national lead- ers to cede power from the center to subnational levels. Csergö, Roseberry, and Wolff (2017) suggest that subnational governance arrange- ments in which the central government cedes political or economic power to subna- tional governance units are only likely in the context of state weakness or if there is a high level of external pressure to do so. Central governments are only willing to devolve genuine power if the cost of doing otherwise is too high. As central state authorities seek to mitigate the cost of ongoing intrastate conflict, subnational governance is stra- tegically applied to curb such costs. Therefore, observations suggest a positive correla- tion between forms of subnational governance and the occurrence of intrastate conflict (Tranchant 2016). Otherwise, the central government is unwilling to grant greater political autonomy or to transfer state resources away from the center and to the periphery (Aalen and Muriaas 2017). It is expected that majorities will be inclined to uphold their commitments to devolve power once the power balance has shifted (Mogaka 2017). When power bal- ances change in a conflict’s immediate aftermath, the stronger party has incentives to renege on its commitment and to demand better terms (Walter 2009; Powell 2006). Lake and Rothchild (2005) concurred that subnational governance arrangements in the form of territorial decentralization were unsuccessful in the long term due to bar- gaining failures between majority and minority groups. This is the case because “even though a majority at the center might be committed to decentralization and willing to pay the cost involved today, this commitment can be—and, if the historical record is any judge, is likely to be—overturned in the long run” (Lake and Rothchild 2005, 129). The changing power dynamics between the center and the periphery, and between majority or dominant groups and minority or marginalized ones, critically influences the success of subnational governance reform as a conflict mitigation mechanism. Gaynor (2016) found that resistance from the central state authority had proven to be a major obstacle in implementing political settlements with subnational governance arrangements. Meanwhile, rural elites have an incentive to reinforce their domination over local communities (Boone 2003). Non-state armed actors that hold territory are often unwilling to enter into an agreement with warring factions, despite the potential for stability and economic prosperity. This is because they are concerned about not receiving a return on investment after relinquishing power (World Bank 2017). Negotiating parties and mediators have to consider the inclusion of formal dispute settlement mechanisms to guide the implementation of the political settlement and to 72 Subnational Governance and Conflict resolve disputes between the central government and subnational units. Such mecha- nisms encourage consensus building and may prove important in preventing a flare-up of renewed tensions. Violent conflict is the outcome of the institutional failure to man- age peaceful dispute resolutions (World Bank 2017) and to strike a bargain that takes account of parties’ commitment problems. Thus, power-sharing and power-dividing arrangements have to address credible commitment problems by offering the parties a share of political power and security guarantees in the form of checks and balances (Hartzell and Hoddie 2007). From this perspective, what looks like unwillingness on the part of political elites is in many cases a limit of their political capacity to make concessions to adversaries that their constituencies, in particular hard-liners, do not support, as has been the case in Mali and the Philippines. What Role for Third Parties? Scholars have found that third-party enforcers, such as UN peacekeeping missions, are critical for influencing the various parties’ commitment to peace (Doyle and Sambanis 2000; Walter 1997, 2002; Hartzell and Hoddie 2007). The presence of external actors serves as a guarantor to former adversaries that security is provided and parties are not able to renege on their commitments. Third-party intervention also has an important signaling role. When warring parties reach an agreement to invite and accept the pres- ence of third parties to support and monitor the implementation of a new political settlement, parties let former adversaries know that they are committed to the agree- ments reached in the settlement (Walter 1999). However, for third-party interventions to successfully guide the implementation process, their support has to be perceived as sustainable and long term to foster confi- dence among previously warring parties. The abrupt withdrawal of the U.S. from Somalia in 1993 following an attack and the death of 18 of its armed forces personnel represents a negative example of sustained commitment by a neutral external actor. However, Hartzell and Hoddie (2007) caution that external third parties have to act as a mediator between the former adversaries, rather than an enforcer of the agreement, to mitigate the risk of escalating tensions after third-party forces withdraw. Implementation Is Key The implementation process—not only whether new political settlement provisions have been implemented but also how they have been implemented—is a key factor in determining the success of subnational governance arrangements as a conflict mitiga- tion mechanism. Lessons learned from implementation processes of decentralization in fragile and conflict-affected states reveal that the incomplete transfer of political, fiscal, and administrative powers from the central level to regional and local gover- nance levels, as well as the central government’s reluctance to provide sustained sup- port to subnational governance units, are critical impediments to the success of subnational governance arrangements (Narang Suri 2016). In Yemen, for over two decades there were several attempts to strengthen subna- tional governance, as has been initiated with the passing of the “Local Authority Bill of Framing Subnational Governance to Secure Peace 73 2000” and subsequent regulatory frameworks broadening the mandates of local gov- ernments. Some progress has been achieved in recent years. The legislation provided the legal framework for decentralized local governance structures and introduced elec- tions for local councils, empowering communities. These councils have been able to maintain a sense of security and order and have improved access to livelihoods across many areas of the country that have remained free of conflict. However, uncertainties persist over financial allocations and resource ownership for localities, which in turn have hampered the delivery of services. The failure to implement subnational governance arrangements not only risks the outbreak of renewed violence and intrastate conflict; unsuccessful implementation also has an enormous negative impact on the prospects for future settlement processes. As political settlements collapse, warring parties are less likely to reenter negotiations or to hold up their commitments under the proposed framework (Hartzell and Hod- die 2007). CASE STUDY 7  Mali—A History of Unmet Expectations In Mali, a failure to implement the subnational governance arrangements foreseen in a number of political settlements reached between the central government and separatist movements in the North resulted in an escalation of intrastate conflict and revived demands for independence. Five settlements—the Tamanrasset Accord (1991), the National Pact (1992), the Algiers Accord (2006), the Ouagadougou Accord (2013), and the Algiers Peace Accord (2015) —all contained significant commitments to subnational governance, but their implementation by the central government has been slow and failed to achieve many of its stated objectives of promoting state legitimacy, stability, cohesion, and more efficient service delivery. Put differently, incomplete implementation meant that only one of the underlying conflict drivers was addressed—political exclusion was mitigated by a more meaningful political decentralization process of establishing elected subnational governments. However, little was achieved in terms of reducing economic underdevelopment and the social and cultural marginalization of people living in the North. While most of the earlier conflicts remained localized, the most recent rebellion destabilized the entire country. A failure to implement political settlements, together with social and economic grievances and growing insecurity across North Africa, eventually led to an insurgency in early 2012. The National Movement for the Liberation of Azawad (MNLA), a separatist movement primarily made up of Tuareg people, had taken control of northern Mali, calling again for greater autonomy or independence for a region it calls Azawad. The rebellion in the North and a simultaneous military coup in Bamako resulted in a full occupation of the northern regions by rebel and terrorist groups, which also threatened to overrun Bamako. A foreign military intervention by France in January 2013 brought about a period of stability and a return of democratic governance. However, the conflict resumed in May 2014, when armed groups reclaimed control (Box continued next page) 74 Subnational Governance and Conflict CASE STUDY 7  Mali—A History of Unmet Expectations  (continued) over large parts of the North. Peace negotiations were relaunched in Algiers and culminated in the signing of the Algiers Peace Accord in June 2015 by the national government and two coalitions representing northern separatist movements. The agreement again included provisions for greater autonomy for the North. However, local realities have yet to change for marginalized communities in the North. Several factors triggered Mali’s instability and near collapse in 2012, including regional events linked to Libya’s revolution and the growing influence of criminal networks and terrorist groups in the Sahel. However, the core of the crisis involved long-standing challenges related to Mali’s weak governance structures and the insufficient implementation of devolution foreseen in previous peace agreements. In short, Mali’s efforts in building a democratic political system, reflected by peaceful democratic transfers of power during the past 20 years, are undeniable but have yet to translate into levels of effective and accountable governance that would fully and simultaneously satisfy domestic challengers and external development partners. The policy of decentralization that went hand in hand with democracy was never fully implemented, creating a system of overlapping central, decentralized, and deconcentrated entities that left local governments largely incapable of discharging their functions. Compounded by challenges facing the central government in delivering services in the vast territory, as well as the low density of the North, relationships between the state and society were undermined, creating opportunities for non-state actors at the local level to establish parallel governance structures that further weakened state legitimacy, stability, and social cohesion. Fragility and conflict in Mali is reinforced by the central government’s limited presence across the entire territory and weak institutional capacity, which eroded its legitimacy, undermined the social contract, and rendered it unable to exert state authority or defend the country’s territorial integrity. For instance, in the period between December 2019 and March 2020, only about 23 percent of civil servants were present at their duty stations in northern Mali.a As a result, Mali’s conflict is driven by systemic institutional weaknesses related to poverty, real and perceived inequality, and social structures, which previous governments have not addressed in a sustainable way. In 2010, over 50 percent of the population lived in poverty, and 90 percent of poor people lived in rural areas. Growing social tensions are reinforced by the presence of criminal networks and terrorist groups, foreign influences, climate change, and population growth, which is increasing competition over scarce opportunities to secure livelihoods. Subnational governance has been a component of the Malian state structure since independence in 1960, but it became more pronounced when the democratization process was launched in the early 1990s. The subnational governance arrangements outlined in the constitution sought to address the grievances of Tuareg communities in the North, and represented a conscious effort to distribute power to lower levels of government. The 1991 Tamanrasset Accord recognized a special status for the three northern regions of Kidal, Gao, and Timbuktu, and promised to increase development-related efforts in these parts of the country. In the 1992 National Pact, the central government again agreed (Box continued next page) Framing Subnational Governance to Secure Peace 75 CASE STUDY 7  Mali—A History of Unmet Expectations  (continued) to an administrative reorganization of the territory. Three levels of subnational governance were proposed—interregional, regional, and local—as well as directly elected regional assemblies. However, an asymmetrical subnational governance arrangement was never implemented. The central government instead established a unitary state system with high levels of decentralization for all regions. Three subnational levels— collectivités territoriales —were designed, including regions, cercles, and communes, as well as the special district of Bamako. Contrary to promises in the National Pact that the regions would become the primary level of subnational governance, the decentralization law assigned the communes most governing authorities at the local level. A flare-up of violence in the early 2000s compelled a renewed commitment to the promises made in the National Pact. The 2006 Algiers Accord announced the establishment of a temporary regional council for coordination and a special investment fund to develop the North. It also established a degree of autonomy on security matters by forming special military units. The severity of the 2012 crisis revived the debate about state presence in the North and the roles and responsibilities of decentralized governance. The 2015 Algiers Peace Accord constituted a new attempt to address the challenges of governance and local representation. In particular, it included commitments to increase northern populations’ representation in state institutions and to reform the existing High Council of Local Authorities to ensure representation of traditional authorities, youth, and women. It also sanctioned regionalization as a guiding principle for further decentralization in Mali, as reflected in strengthening the presidency of the Regional Council, who will serve as chief executive in the region. The agreement also proposed to establish two new regions in the North and provided for a review of the administrative supervision exercised by the central government over the local authorities. Recent survey results suggest that only 17 percent of the population in the North is confident that the central government will be able to implement the 2015 Algiers Peace Accord. However, there is a sense that the severity of the 2012 crisis will result in a real and tangible commitment to move beyond rhetoric and facilitate implementation of key decentralization reforms. Additionally, an Agreement Monitoring Committee has been established to monitor and guide the implementation of commitments made by the various parties. Thus far, implementation of the different elements of the political settlement reached in 2015 has been delayed. The proposed interim government that was tasked with extending state authority in rural areas was not installed until January 2017 and was unable to establish a state presence in large parts of the country’s territory due to an escalation of violence in northern and central Mali. In September 2017, legislation adopted by the National Assembly on the parameters for self-governance and on the implementation of the new code des collectivités territoriales was rejected by a coalition of northern separatist movements and subsequently reopened for revision. Subnational elections for interim authorities held in 2016 lacked credibility and were accompanied by political instability, and (Box continued next page) 76 Subnational Governance and Conflict CASE STUDY 7  Mali—A History of Unmet Expectations  (continued) a coalition representing Tuareg separatists, including the MNLA, boycotted the elections. Regional and local elections aimed at replacing the interim measures were scheduled for December 2017 and April 2018, respectively, but they were continuously postponed. After parliamentary elections were also postponed several times citing security concerns, eventually, on 19 April, polls closed for the second round of parliamentary elections, the first held since 2013. The elections are considered an important milestone in advancing the implementation of the 2015 Algiers Peace Accord as parliamentarians are expected to adopt amendments to the constitution that would provide the framework for decentralization. To hold the legislative elections before May has been a commitment emerging from the national dialogue that was held in several sessions throughout 2019 concluding in December and also resulting in agreements on a constitutional referendum, the re-establishment of state presence throughout the country, and a review of the 2015 Algiers Peace Accord. Despite an agreement on a revised “political road map” reached between the main parties in March 2018, intercommunal violence flared up ahead of presidential elections held in July 2018. A report issued by the UN in August 2018 found that delays of institutional reforms concerning decentralization “provided armed groups with an argument to postpone their own commitments on security matters.” In response to the deteriorating security situation and a potential collapse of the peace process, parties to the 2015 Algiers Peace Accord signed a Pact for Peace on October 15, 2018, in which they recommitted to implementing the agreement. Only days before, on October 10, 2018, a draft legislation on the demarcation of subnational units was leaked and caused upheaval; the proposed division of subnational units into 20 instead of the current 10 regions, and the establishment of new cercles was viewed as favorable to nomadic groups, in particular the Tuareg community.b In 2016, 11 cercles and the two regions of Taoudenni and Menaka were established in the North. The lack of progress and insufficient transfers of resources and responsibilities resulted in increased frustration about slow implementation and lack of accountability. While local elections and autonomy were intended to strengthen state-society relations, the central state maintained a large degree of top-down authority, despite the fact that local governments were independently elected and had expenditure responsibilities devolved to them. Subnational governance reform in Mali was based on a significant transfer of responsibilities and resources to local governments. The three levels of collectivités territoriales were nominally handed responsibility for a substantial portion of public service provision in several areas. Key devolved competences included health, education, and infrastructure. Fiscal decentralization has lagged behind, since responsibilities have been transferred without accompanying resources, and institutional decentralization retains a significant role for the state’s exercise of its supervisory capacity. The 2015 Algiers Peace Accord foresees that the central government transfers 30 percent of public revenues to the subnational units. In the 2019 budget, the government earmarked 23.1 percent of public revenues for the transfer to (Box continued next page) Framing Subnational Governance to Secure Peace 77 CASE STUDY 7  Mali—A History of Unmet Expectations  (continued) subnational authorities. However, the absence of an adequate banking system in the North meant that the government was unable to complete the transfer of these allocations. Moreover, in July 2019, legislation was adopted for the creation of a ‘Northern Development Zone’ to provide the framework for the operationalization of development projects in the North.c According to the 1995 General Code on Territorial Collectivities in the Republic of Mali (Loi n°95‑034 du 12/04/1995), local governments, and in particular the communes, should have their own resources and autonomous budgets, which should grant them the necessary means to fulfill those responsibilities transferred by the central state. Intergovernmental transfers, however, are not nearly sufficient to fill the gap in local government budgets. This system of dependency on the state helped tilt accountability of locally elected officials away from the local population toward the central government, thus undermining the very basis of democracy that it was intended to promote. The government has made great strides in seeking to advance the implementation of the 2015 Algiers Peace Accord. Yet, the UN Secretary-General concluded in his quarterly report on Mali of December 2019 that “[he is] concerned about the continued delays in the implementation of the agreement.”d The 2012 crisis brought to the fore questions about whether subnational governance and local elections actually promote peace and territorial integrity. Beyond questions of fiscal viability of the system as a whole, the northern regions’ cohesion challenges seem to have been left unaddressed by the different waves of subnational governance reform. Demarcations drawn along socio-politico- economic delineations in the North perpetuated a sense of ethnic and subnational identity as the basis for political organization in which state penetration was weaker. While a prevailing perception of marginalization among northern groups continues to fuel instability, the largest share of the country’s poor reside in southern regions. This tension—between reality and a perceived marginalization— suggests that grievances might not be addressed through renewed investments in service delivery but rather that the tension itself should be the focus of policy interventions. This adds to the magnitude of the challenges Mali faces and highlights the need to address grievances comprehensively rather than in a piecemeal fashion. Notes: a. United Nations Document, 2020. b. International Crisis Group 2018b. c. United Nations Document, December 2019. d. United Nations Document, December 2019, p. 15. 78 Subnational Governance and Conflict CHAPTER 4 Strengthening Subnational Governance in Political Settlements As noted in Chapter 2, intrastate conflicts—including those in which control over terri- tory is a major source of contestation—are frequent, destructive, and long lasting. At the same time, consensus over the causes of and facilitating conditions for conflict onset, duration, intensity, and recurrence has been slow to emerge, at best. The debate over which approaches to conflict mitigation are effective is also far from fully conclusive. While we cannot resolve either of these perennial debates, our study seeks to make a con- tribution by focusing on countries affected by fragility, conflict, and violence in which subnational governance arrangements have been an integral part of a conflict mitigation strategy resulting in a new, engineered political settlement. This study aims to identify cases that represent the wide spectrum of variation across different dimensions, such as geography and demography, and the nature of both the conflict and of international involvement, to reflect in breadth and depth about the conditions under which viable subnational governance structures can emerge as part of sustainable political settlements. Of the seven case studies chosen, Bosnia and Herzegovina represents a historical case that offers a great source for lessons learned. It demonstrates the potential role external actors can play, including the World Bank. Kenya highlights how subnational governance reforms that promote the redistribution of public resources reduce the intensity of competition among ethnic groups for control of the central government. In Mali, the main ethnic minority group is territorially concentrated, but its claims for greater autonomy have merely been met on paper. Therefore, Mali demonstrates the consequences of failing to implement past peace agreements: subnational (and other) governance provisions become meaningless, and the incentives for peripheral groups to engage in peace processes significantly reduce the prospects of political stability and economic development. Myanmar’s conflict has taken place in a context characterized by its long duration—more than seven decades—and its highly fractionalized, plural society. While the introduction of subnational governance arrangements has generally been recognized as a useful conflict-mitigating strategy, the complexity of the 79 country’s postauthoritarian transition compounds the difficulty of parallel efforts to initiate a postconflict transition, and vice versa. In the Philippines, a similarly long process of conflict mitigation qua subnational governance arrangements has become increasingly focused on asymmetric fiscal decentralization to address the enormous economic inequality faced by territorially concentrated Muslim minority groups. Somalia, in turn, represents a different challenge of asymmetry. This case clearly illus- trates the dual complexity of building a federal governance system in the historical absence of a legitimate and functioning central government and in the presence of regions that have varying degrees of autonomy and capacity. This already complex domestic setting is further complicated by the challenge of external involvement in an ongoing conflict. Against the backdrop of persistent instability in the Middle East, Yemen illustrates the difficulties of managing meaningful subnational governance reform triggered by the Arab Spring. Here, the principle of subnational governance had been accepted, but how it was to be implemented became a major source of con- flict. Indeed, implementation pitted different domestic factions and their equally rival- rous external backers against each other, making a negotiated transition from dictatorship to a more open political system impossible. Reflecting on these diverse experiences, the case studies consider a number of ques- tions that, in turn, reveal some more generalizable lessons on the utility of subnational governance structures as part of international conflict mitigation efforts. As noted earlier, these questions include: Which conflict causes could have been/were addressed by subna- tional governance arrangements? Were these arrangements correctly identified, or did they miss the target? What other governance arrangements formed part of the negotiated agreement? To what extent and why were these arrangements implemented and sustained or revised after implementation? Which were the most pivotal factors, and who were the most pivotal actors in explaining this outcome? On this basis, the section offers a brief summary of all seven cases, focusing on the critical lessons that can be learned from them with respect to the role of subnational governance arrangements in negotiating, imple- menting, and operating a new political settlement. As such, this section synthesizes the relevant lessons that can be applied to the future engagement of international develop- ment partners with countries recovering from fragility, conflict, and violence. Three Conclusions: Process, Content, and Context Any institutional arrangements put in place to settle an intrastate conflict are likely to play a conflict mitigating role only to the extent that they (directly) address conflict parties’ motivations and make conflict less feasible. Each case study offers important insights into the conditions under which conflict mitigation is possible. This section synthesizes these insights and offers conclusions on lessons about which drivers of conflict are likely to respond to what kinds of subnational governance arrangements and under what conditions, and how development partners can help conflict parties negotiate, implement, and operate such arrangements in the long term. Among the critical lessons is that political settlement processes are always highly politicized domestically, since they reflect, and then institutionally reshape, power bal- ances between different political actors at the beginning of a postconflict transition. Development partners involved in such processes must be aware of this and 80 Subnational Governance and Conflict understand that their engagement, however neutral it may have been conceived, has political impacts that affect the way in which they are perceived by domestic actors and that can increase or limit the effectiveness of their role in any postconflict transition. At its very essence, subnational governance is concerned with the distribution of political power. When governance arrangements seek to shift power away from the center to lower levels of governance, there will always be winners and losers among the political elite. Consequently, each process of engineering a new political settlement that involves subnational governance arrangements will engender various degrees of opposition and support for a change in the status quo. Regardless of how important the role of development partners and other interna- tional actors might be in initiating a transition from war to peace, a sustainable peace is one that is owned locally and as inclusive as possible of all conflict parties, while also reflecting actual power balances between them. Political settlements require active implementation; they hardly ever self-execute. International development actors need to play their part in avoiding implementation delays and failures, including providing security guarantees and helping (with technical expertise and financial resources) build local capacity to operate the institutions agreed on in a settlement. Beyond this relatively generic framing, it is obvious that the success of subnational governance arrangements in contributing to sustainable peace, democracy, and devel- opment in the aftermath of violent intrastate conflict depends on a number of factors and varies, at times significantly, across cases. Yet it is nonetheless possible to identify several critical and more general lessons from the case studies across three dimensions: the content of the institutional package, and the process and context in which it is negotiated, implemented, and operated. There is a certain and unavoidable degree of overlap between these three broader dimensions and the individual factors in each of them, as shown in Table 2. TABLE 2  The Process, Content, and Context of Political Settlements Process Inclusiveness Conclusiveness Content Match of institutional arrangements to conflict drivers Appropriate dispute-resolution mechanisms Constitutional entrenchment and other guarantees Sufficient and sustainable financing Context Domestic Parties’ acknowledgment of each other’s right to be part of the common state with their distinct identities Parties’ commitment to good-faith negotiation, implementation, and operation Absence of effective domestic spoilers Matching institutional bargaining to the effective balance of power External Absence of effective external spoilers International support for design, implementation, and operation Strengthening Subnational Governance in Political Settlements 81 Process Table 2 identifies two dimensions of process in relation to the institutional packages negotiated as part of engineered political settlements: their inclusiveness and conclusiveness. Inclusiveness Inclusiveness relates to the participation of all relevant interest groups in the negotia- tions, at the end of which a new institutional structure is agreed. The underlying logic here, in line with the political settlements perspective, is that unless all powerful actors are included and have an opportunity to shape future institutional arrangements, any negotiated outcome is unlikely to adequately reflect the effective balance of power. A lack of inclusiveness, consequently, will not result in a political settlement that can produce the required political stability to be sustained, since it will produce effective spoilers. In Myanmar, the current constitution’s legal clauses still secure a politically domi- nant position for the military. Grievances related to the political and economic mar- ginalization of ethnic minority groups, as well as those regarding unjust resource management, remain unresolved. Recent incidents in Rakhine State also demonstrate the lasting, deep-rooted, religious divide between Buddhist and non-Buddhist com- munities. The peace negotiations excluded a number of opposition groups that were not recognized by the government. The October 2015 Nationwide Ceasefire Agree- ment (NCA) was only signed by eight opposition groups, with a further two signing in 2018. However, groups that have not signed control more than twice as many troops as those that have signed. The lack of progress on generating substantive agreements on what federalism should look like is among the main reasons why the peace process has stagnated. Conclusiveness Conclusiveness refers to the need to address all conflict-critical issues in the negotia- tion process. This need not imply that they all be resolved as part of negotiations, but they must be recognized as issues to be addressed as part of the broader process of engineering a new political settlement. Some issues may need to be postponed, while others may need to be delegated to specific bodies, either to work out a technical (rather than political) solution or to arbitrate in a particular dispute. The rationale for postponing or delegating specific issues is to achieve the maximum possible deal, not to endanger such a partial political settlement by striving for a comprehensive yet unattainable one. As a partial deal is secured and proves viable—by providing political stability and generating sufficient economic benefits for participating elites and their constituents—subsequent compromise on unresolved issues is more likely. The long history of Yemen’s North-South division is a central factor in its ongoing political crisis. While the peace negotiations during the National Dialogue Conference (NDC) sought to adequately address the grievances of southern constituents, they were not able to resolve the conflict-critical issues. Since the South gained 82 Subnational Governance and Conflict independence from British rule in 1967, the relationship between the two regions has perpetually alternated between hostility and partial accommodation. The two coun- tries eventually unified in 1990 under the Saleh presidency. However, beginning in the early 2000s, President Saleh faced growing alienation from the South, and toward the end of his regime, he had to deal with a strong separatist movement in that region. Given the recent expansion of the Houthi movement southward, calls for indepen- dence in the South are growing louder. With the support of President Hadi, who con- sidered reconciliation with the South a political priority and emphasized an inclusive political process to lead the way for a new constitution, there has been weak consensus in favor of adopting federalism as a state structure for Yemen. The NDC included a working group on the “southern issue” and proposed federal- ism and a special power-sharing arrangement as mechanisms to mitigate secessionist tendencies in the South. While consensus was reached in favor of federalism, after 30 meetings the NDC has been unable to develop a governance structure that addressed disputes over the boundaries and numbers of regions. The subnational governance arrangements outlined in the draft constitution have constituted Yemen with six regions, four in the North and only two in the South, further alienating marginalized constituents there. The proposal was rejected by opposition groups from the South on the grounds that Yemen should be built on a two-unit federation, along North-South boundaries based on a partnership of two equals. Representatives of the North, fearing that a two-unit federation might prove unstable and eventually lead to secession, have dismissed this proposition. Content As illustrated in Table 2, there are four crucial elements concerning the content of any agreement over the design of the institutions, processes, and mechanisms that form the institutional package for resolving differences between the parties within a new political settlement. Drawing on findings from the case studies, this section discusses these four elements in turn. Match of Institutional Arrangements to Conflict Drivers As noted in Chapter 2, not all conflict drivers are equally actionable, and not all drivers are equally responsive to conflict mitigation qua subnational governance arrange- ments (or, for that matter, other institutions such as power sharing or wealth sharing). Thus, it is important to look beyond what conflict parties want and establish whether their desires address particular conflict drivers, and do so in a genuinely conflict- mitigating way that does not exacerbate the effects of other conflict drivers. In addi- tion, it is critical that there is a truly comprehensive settlement—that is, one that deals with all identified drivers, which in turn also requires that institutions fit with each other as a “package,” and considers possible tensions between a traditional governance system already in place and new, formal arrangements being introduced as part of a political settlement. In Kenya, conflict was driven by political and economic exclusion and disputes regarding control over political power and state resources between groups who Strengthening Subnational Governance in Political Settlements 83 identified along ethnic lines. Subnational governance arrangements have been applied to address these conflict drivers. In response to the 2007–08 postelection violence, Kenya’s political leaders reached a power-sharing agreement in 2008 and launched an extensive devolution program in 2010, seeking to prevent the recurrence of postelec- tion violence by reducing the stakes in winning (or losing) the office of the presidency. Kenya’s winner-takes-all electoral system exacerbated tensions in its highly fragmented society, where resource allocation is strongly determined by tribal belongings. The 2008 power-sharing agreement established a coalition government between the Party of National Unity (PNU), representing the Kikuyu ethnic group and headed by Presi- dent Mwai Kibaki, and the Orange Democratic Movement (ODM), with its support base among the Luo and led by Raila Odinga. To safeguard against the possibility of the opposition group reneging on its commitment, the position of prime minister was created as a transitional measure that was not incorporated into the new constitution. The 2010 Constitution established a system of devolved governance dispersing politi- cal power and economic resources away from the center and toward lower-level coun- try governments. Appropriate Dispute-Resolution Mechanisms Engineered political settlements are normally characterized by a degree of (construc- tive) ambiguity in their provisions. This allows negotiators to claim that they achieved what they promised their constituencies and to keep them on board during the imple- mentation process. In addition, political settlements are often negotiated under the pressure of time, the danger of collapsing ceasefires, or the defection of one or more conflict parties. Individually and collectively, these factors often create situations in which disputes arise over the interpretation of particular provisions during a settle- ment’s implementation phase. Disputes may also arise later on, for example, between authorities at different levels of government, such as a subnational and a central gov- ernment. In order for these disputes not to derail a peace process or destroy a political settlement, it is important that parties agree to dispute resolution mechanisms as part of a political settlement that will subsequently enable them to resolve differences by political means, rather than by renewing their armed conflict. These can take various forms, from special commissions to judicial review, and may include international representatives. Somalia needs to improve its institutional capacity for successful dispute resolu- tion. With access to political power and state resources at the national and regional levels at the heart of disputes between political actors and local groups, state building and questions regarding the divisions of political power and state resources are inher- ently contentious issues. Subnational governance arrangements as proposed by the 2012 Provisional Constitution did not adequately reflect realities on the ground, par- ticularly as they relate to subnational boundaries and their interference with existing informal structures. The central government seeks to strengthen de facto authority in most of the country outside the capital of Mogadishu as a way to improve its capacity to enforce agreements on the formation of federal member states, provide dispute res- olution, or guarantee security. Historically, in the absence of formal governance, 84 Subnational Governance and Conflict traditional and informal governance structures have served as dispute resolution mechanisms at the local level, particularly in Somaliland and Puntland. Constitutional Entrenchment and Other Guarantees Constitutional entrenchment is a critical guarantee for a political settlement’s sustain- ability. Constitutional and other guarantees send a clear signal to all parties and their constituents about the commitment they are making to implement and uphold the settlement, while leaving open the possibility for future revisions by broad, and nor- mally mutual, consensus. Committing to building guarantees into formal legal struc- tures and informal institutions also creates incentives for constructive engagement in negotiations because it makes it more likely that any institutional arrangements agreed on will in fact be subsequently implemented and operated. In Bosnia and Herzegovina, this dimension was closely intertwined with the inter- national community’s strong involvement in the conflict-termination process. The country’s new constitution became part of the Dayton Peace Accords, which are them- selves an international treaty and were endorsed by the UN Security Council. The Day- ton Accords also mandated a UN Protection Force and an International Police Task Force, thus simultaneously providing security guarantees. Dayton also established a range of international oversight bodies, such as the Office of the High Representative, and provided for international judges in the country’s constitutional court. Constitu- tional entrenchment in Bosnia and Herzegovina goes hand in hand with power- sharing provisions that require various qualified and concurrent majorities for any amendments to the Dayton Peace Accords. Sufficient and Sustainable Financing Political settlements remain commitments on paper until they are implemented. While implementation is first and foremost a question of conflict parties’ political will, it is also dependent on the availability of resources. Especially in cases involving subna- tional governance arrangements, it is critical to ensure that resources are available for these institutions to be created (where they do not already exist) and to be sustainably financed so they can deliver public goods and services. In addition, it is also critical that financing mechanisms are established such that old inequalities are not exacer- bated, or new ones created. Therefore, the availability of revenue sources must match the subnational governance arrangements, particularly as countries that had been held captive by violent conflict for years and sometimes decades lack fiscal resources and institutional capacity. In the Philippines, the peace agreement initially awarded the Mindanao region with significant fiscal autonomy without ensuring that the necessary conditions to meet these responsibilities were put in place. Fiscal responsibilities were devolved to the regional level but without the authority to raise government revenues. This forced regions to rely on intergovernmental fiscal transfers from the center, which created a situation in which the regional governments were not accountable to the central gov- ernment or to the local constituency. In response to lasting economic grievances and disputes over resources, the Bangsamoro peace agreement emphasized fiscal Strengthening Subnational Governance in Political Settlements 85 autonomy through revenue sharing and the creation of intergovernmental fiscal institutions. Context The institutions, processes, and mechanisms that are part and parcel of any political settlement do not operate in a vacuum, but in what is generally a highly politicized and volatile environment, characterized by high levels of distrust among the presumptive parties to a political settlement and the very real dangers of the presence, or emer- gence, of spoilers attempting to derail a settlement process. While settlement design (i.e., the content dimension of success) can address some of these issues, both the domestic and external contexts in which a settlement is designed, implemented, and operated require active management, including by international development partners. Parties’ Acknowledgment of Each Other’s Right to Be Part of the Common State with Their Distinct Identities Unless they are explicitly designed as interim or transitional settlements, political set- tlements need to be based on a minimum consensus between parties regarding each other’s right to be part of a common state with their distinct identities. In the context of conflicts with subnational dimensions, this cuts both ways: it requires an end to secessionist activities and the genuine accommodation of difference, including through the creation of spheres of self-governance for distinct groups at the subnational level. In Myanmar, ethnic minority communities and armed groups had been struggling for greater autonomy for over six decades and advocated for a return to the “spirit of Panglong,” which promised them devolved authority over their regions. They consid- ered that this promise was broken in 1962 when General Ne Win abrogated the 1947 Constitution and abolished the ethnic councils set up under that constitution, thereby dissolving local governments and bringing all areas of the country under centralized and uniform administrative control. This remains the case under the 2008 Constitu- tion, which, although it provided for some decentralization of power, still maintained a fairly centralist conception of the state. It established some form of subnational gov- ernance to the 14 states and regions and the 6 self-administered areas. However, the degree of self-governance remains limited; the legislatures of these decentralized units, though locally elected, have no exclusive lawmaking powers. Similarly, the decentral- ized executive powers are minimal, and officeholders are not locally accountable, since the chief ministers of the states and regions are appointed by the president. Parties’ Efforts toward Implementation Implementation is naturally a critical component of a political settlement’s success. It requires joint efforts by parties at the central level, including managing a constitu- tional reform process, as well as individual efforts, including delivering governance at a subnational level in line with the provisions of a political settlement and/or an over- all constitutional order. With the exception of Bosnia and Herzegovina, which had just 86 Subnational Governance and Conflict declared independence before conflict broke out, all countries analyzed in this study have had some form of legal framework that constituted a governance structure with elements of subnational governance. Yet a failure to implement, or the lack of feasibil- ity of these subnational governance provisions, has meant that deep-rooted grievances have remained unaddressed and resulted in renewed violence. Next to long-term governance reform and institution building, affected communi- ties need to benefit from quick peace dividends following the signing of a peace agree- ment. Political settlements raise citizens’ expectations for positive change. Yet when seeking to support the central government in rebuilding core governance functions, development actors must be realistic about the government’s capacity to absorb for- eign aid and use it to effectively implement the institutional aspects of an engineered political settlement. This further exacerbates the difficulties of sustaining a new politi- cal settlement because in the immediate aftermath of conflict, there is normally a lack of institutional infrastructure that could effectively translate the availability of devel- opment funding into public service delivery. Absence of Effective Domestic Spoilers Because they are compromises, political settlements hardly ever satisfy all conflict par- ties. However, the more balanced the compromises, and the more inclusive the negoti- ation processes, implementation, and operation of the resulting settlements, the less likely it is that effective spoilers with the (military) capacity to reignite major armed conflict will emerge. In Yemen, any political settlement has always been fluid. Shifts in the power balance among elites were frequent and caused by changes in political alignments. Reflecting such changes, relations between subnational entities, and between them and the cen- ter, were also in constant flux. This domestic situation, in turn, was subject to the influ- ence of a variety of foreign actors and also shaped the regional dynamics of security and stability. Former President Saleh has acted as a spoiler to the NDC and the transi- tion process. After having been pushed out of power through internal and external pressures, Saleh built a new power base and managed to derail the state-building pro- cess, which led to the devastating civil war that is ongoing. Saleh has taken advantage of his sustained influence on his political allies and previous role in the ruling General People’s Congress and that party’s financial resources, bureaucratic experience, and historically powerful position. Equally important, however, is the fact that demands by Houthis were not sufficiently reflected in the subnational arrangements proposed by the NDC, which allowed spoilers, including former President Saleh, to exploit their grievances and mobilize against the central government. This created a powerful mili- tary alliance that fatally undermined the country’s transition process and its new polit- ical settlement. Absence of Effective External Spoilers Intrastate conflicts frequently draw in external parties in support of government or rebel factions, or sometimes in pursuit of their own interests. Settlement negotiations, too, tend to involve a variety of external actors, often with conflicting agendas. This Strengthening Subnational Governance in Political Settlements 87 creates the possibility of external state and non-state actors becoming spoilers in polit- ical settlement processes on their own or in cooperation with allies (or proxies) in the domestic context. Limiting the likelihood of such spoilers to emerge or to be effective is thus an important element of successfully managing a political settlement. Active interference of external actors also affects, often negatively, local constituencies’ sense of national ownership over the peace process. Both of these negative externalities have been at play in Somalia. Active engagement by the international community—including the African Union, the UN, and the World Bank—to drive the peace process has left Somalis feeling marginalized and expressing a lack of ownership and participation in political decision making. Interference by neighboring countries, particularly Ethiopia and Kenya, hindered real progress on achieving peace and stability. The international community was perceived as lacking neutrality. With foreign aid having represented a critical factor driving competition among political actors, development actors have to make sure they do not inadver- tently fuel instability. Similarly, Ethiopia, backed by Western donors, had always pushed for a decentralized system of governance in Somalia, which some factions have viewed as an attempt to keep the country fragmented. Ethiopia’s involvement in Somalia was linked to the territorial dispute over the Somali-inhabited Ogaden region and its con- flict with Eritrea, which in turn sought to exploit Somali hostility toward Ethiopia following the 2006 intervention. Furthermore, different regions were allying with dif- ferent external actors supporting their claims, such as Puntland and Somaliland with Ethiopia, and Jubaland with Kenya. In contrast, Arab countries have expressed support for a strong centralized state. The competing regional interests even led to the facilita- tion of two rival peace ­ conferences—in Sodere, Ethiopia, in 1996 and in Cairo, Egypt, in 1997—with conflicting state-building agendas. In addition to the presence of domestic spoilers, Yemen’s civil war has become more protracted due to the involvement of Saudi Arabia and Iran. Yemen has become another playing field for the proxy wars between the two states as they vie for regional power and influence along sectarian lines. In 2015, Saudi Arabia launched an interven- tion intended to secure its southern border, claiming that Iran was supplying the self-declared Houthi government with weapons. Since then, the United Arab Emirates, the United States, and a number of countries from Africa and the Middle East joined the Saudi-led coalition. The coalition announced that its efforts sought to support the UN-led peace process; instead, it has caused a devastating humanitarian crisis and pushed the actors into a political deadlock. Matching Institutional Bargaining to the Effective Balance of Power Closely related to the institution-driver match is the issue of whether institutions, in a broader sense, entrench the effective balance of power at the end of the civil war and/or are able to shape a balance of power that can underpin viable state structures in the longer term. This is well illustrated by the case of Bosnia and Herzegovina. The Dayton Peace Accords recognized demographic and geographic realities created by war and sought to organize the country according to a four-tiered governance structure with two 88 Subnational Governance and Conflict entities (the Federation of Bosnia and Herzegovina and the Republika Srpska), ten cantons within the Federation entity, and local communities/municipalities in both entities. While reflecting the status quo at the end of the war and making a new politi- cal settlement possible, this system has also been criticized for its institutional immo- bility and for rewarding, almost in perpetuity, those responsible for unleashing the conflict in the first place. International Support for Design, Implementation, and Operation Given the multiple challenges that political settlements face, countries emerging from often prolonged and destructive intrastate conflicts are unlikely to have the capacity to design, implement, and operate their own political settlements. While local ownership of a deal is critical, in itself it is not sufficient for sustainable peace. However, local commit- ments to a new political settlement can be enhanced by the right kinds of international support, including the provision of security guarantees. As outlined by Walter (2006) and shown by a number of country examples, the perception that the security situation has not improved following peace conferences or the signing of peace agreements creates doubts about the effectiveness of such political exercises. Peace agreements have to out- line credible stabilization mechanisms for the early stages of implementation to address security concerns of weaker parties, in particular minority groups. As noted earlier in relation to guarantees, in Bosnia and Herzegovina, external actors made a strong commitment to provide post-agreement security mechanisms. A 6,000 member-strong NATO peacekeeping force sought to counter the power imbal- ances between local armed forces and to provide third-party security guarantees to secure credible commitment to the 1995 Dayton Peace Agreement. Bosnia and Herze- govina has also significantly benefited from international expertise and financial sup- port in managing and sustaining the complex institutional structure of its 1995 political settlement and providing a minimum of economic benefits to political elites and their constituents. Strengthening Subnational Governance in Political Settlements 89 CHAPTER 5 Conclusion This study has identified a set of key findings relevant for international develop- ment partners supporting subnational governance reform in the context of political settlements in fragile and conflict-affected countries. First, we live in a world with an increasing number of conflicts, the nature of which have changed to be primar- ily fought within state borders and over issues related to governance structures— specifically over the subnational governance design that determines political power sharing and the distribution of national wealth, including the access and exclusion of certain groups. Therefore, efforts to resolve fragility, conflict, and violence must address the internal governance structure and state-society relations that define the nature of the state. Second, each country undergoing a political transition and fac- ing violent conflict engages in a process to reevaluate the governance structure and intergovernmental system. Efforts to design, implement, and ensure the function- ing of such arrangements must consider the underlining drivers of intrastate con- flict and how subnational governance reforms can address these. Our findings reveal that including subnational governance arrangements in politi- cal settlements cannot guarantee sustained peace and that the implementation and functioning of such arrangements bring with them a range of challenges. That said, comparative analysis suggests there are no feasible, alternative governance structures that raise the odds of successfully mitigating conflict. In particular, in countries with plural societies, as well as those engulfed in protracted intrastate conflict with subna- tional dimensions, the combination of subnational governance arrangements with other governance arrangements can be a highly effective tool for fostering lasting sta- bility and prosperity. Every country that transitions from war to peace or from author- itarianism to democracy must reach societal consensus on these fundamental issues. Hence, this study seeks to answer important questions regarding how to maximize the successes and minimize the failures in designing, implementing, and promoting sub- national governance arrangements in political settlements. In light of this, the study concludes that subnational governance is not a panacea for mitigating intrastate conflict. Applying subnational governance as a mechanism to prevent, mitigate, and resolve violent conflict requires a nuanced understanding of the intricacies of subnational governance arrangements and how they play out in different 91 contexts. It recognizes the need to adapt a context-specific approach and that the suc- cess of political settlements with subnational governance arrangements depends on the successful implementation of these provisions. Effective subnational governance can indeed reduce the likelihood of violence and mitigate conflict, whereas poorly designed and incompletely implemented subnational governance reform may exacer- bate tensions and renew intrastate conflict. 92 Subnational Governance and Conflict APPENDIX I Case Studies Matrix 93 94 Country Groups Conflict Governance model Peace agreements Constitutions Bosnia and Bosniaks, Croats, In 1992, Bosnia and Herzegovina Based on the Dayton agreement, November 21, 1995: Dayton Herzegovina and Serbs declared its independence. Bosnia and Herzegovina has a Peace Accords, which arranges However, the Yugoslav National highly decentralized governance Bosnia and Herzegovina into two Army and local Serb militia structure with the framework of entities that are de facto separate challenged its independence a parliamentary representative governments and administrative almost immediately, resulting democracy. The government organs with substantial powers to in a four-year war. In 1993–94, system under Dayton is pass legislation, impose taxes, and hostilities also broke out between consociationalism. otherwise govern. secessionist Croats within Bosnia and Herzegovina, and the army of the Republic of Bosnia and Herzegovina, mainly led by Bosniaks. Kenya Luo, Kikuyu, and In Kenya, the elections of Devolution based on a two-tier February 28, 2008: National Accord 1969: Constitution others 1992, 1997, and 2007 were system composed of the national and Reconciliation Act. 2010: Constitution, characterized by state-sponsored government and 47 county Subnational Governance and Conflict which established county ethnic violence. Violence was governments. administrations based on driven by centralization of political devolution. power in the presidency and inequalities among communities perceived of in ethno-geographic terms. Mali Tuareg Intrastate conflict between Hybrid system based January 6, 1991: Accords of 1960: Constitution the government and separatist on decentralization and Tamanrasset establishing peace 1974: Constitution movements from Northern Mali, deconcentration. Collectivités between Mali and the Tuaregs. which are fighting for greater territoriales are split into 1992: Constitution November 11, 1992: National Pact autonomy or independence for a regions, districts, and giving Expression to the Special region they call Azawad. municipalities. Status of Northern Mali. 2006: Algiers Accords. June 2013: Ouagadougou Accord. June 2015: Accord for Peace and Reconciliation (Algiers Peace Accord). Country Groups Conflict Governance model Peace agreements Constitutions Myanmar Karen, Kachin, Intrastate conflict between the Federalism, but remains October 15, 2015: NCA. 1947: Constitution Karenni, Shan, government and over 20 non- characterized by centralized The government signed the NCA with 1974: Constitution Rakhine, Chin, state armed groups demanding a decision making. Myanmar eight rebel groups, with two more Mon, and others federal state system and greater comprises seven states and 1993: Constitutional convention signing in 2018. The talks focus on autonomy for their regions. seven regions, with six self- 2008: Constitution, based on constitutional revisions determining administered zones or divisions a federal governance system the degree of regional autonomy and and one union territory. granting minority groups a certain political participation. degree of autonomy. Philippines Moros Intrastate conflict between the Local government units elect December 23, 1976: The Tripoli 1973: Constitution central government and armed their own legislatures and Agreement. 1987: Constitution groups from ethnic minority executives. The Bangsamoro January 4, 1987: Jeddah Accord. Moro groups calling for greater agreement lays out the legal autonomy for Mindanao. framework for the autonomous November 7, 1993: Interim GRP-MNLF region of Mindanao as a new Ceasefire Agreement. political entity, with a ministerial July 18, 1996: Final Peace structure of governance, its own Agreement. jurisdiction, and asymmetrical intergovernmental arrangements. May 17, 2007: Resolution No. 2/23- MM on Question of Muslims in Southern Philippines. October 15, 2012: Framework Agreement on the Bangsamoro. Somalia Isaaq and others Civil war driven by interfactional The 2012 Constitution May 18, 1991: Declaration of 1961: Constitution fighting for political power and constitutes Somalia with Independence. 1979: Constitution control over land. federalism, with Puntland and January 8, 1993: Addis Ababa Jubaland declaring autonomy. 2004: Transitional federal charter Agreement concluded at the first The autonomous region of session of the Conference on 2012: Provisional constitution of Somaliland has declared National Reconciliation in Somalia. the Federal Republic of Somalia, independence without which constitutes Somalia with Case Studies Matrix March 20, 2002: Agreement international recognition. a federal governance system, on implementing ceasefire and but political elites fail to agree disarmament, demobilization, and on form and detailed structure of reintegration. federalism. 95 96 Country Groups Conflict Governance model Peace agreements Constitutions Yemen Southerners and Civil war between southern Under the 1991 Constitution, October 28, 1972: Cairo Agreement. 1991: Constitution others separatists and forces loyal to Yemen is a unitary state. Local May 22, 1990: Declaration of 2015: Draft for new constitution the Hadi government on one side, administration is based on Unification. proposed, which has been and Houthis and forces loyal to deconcentrated units of central rejected by the Houthi rebels and former President Saleh on the government departments. March 18, 2013: National Dialogue other conservative parties. other, fighting for control over the Conference, which initiated a central government. Extremist constitution-making process based on violence from ISIS affiliates federal governance arrangements. and Al-Qaeda in the Arabian June 2015: Peace talks in Geneva Subnational Governance and Conflict Peninsula. collapsed. In September 2015, UN officials claimed peace talks would resume. References Aalen, Lovise, and Ragnhild Louise Muriaas. 2017. “Power Calculations and Political Decentral- isation in African Post-Conflict States.” International Political Science Review 38 (1): 56–69. http://journals.sagepub.com/doi/10.1177/0192512115615704 Acemoglu, Daron, and James A. Robinson. 2006. “De Facto Political Power and Institutional Persistence.” American Economic Review 96 (2): 325–30. Bakke, K. M., and E. Wibbels. 2006. “Diversity, Disparity, and Civil Conflict in Federal States.” World Politics 59 (1): 1–50. https://muse.jhu.edu/article/216242 Balch-Lindsay, D., A. J. Enterline, and K. A. Joyce. 2008. “Third-Party Intervention and the Civil War Process.” Journal of Peace Research 45 (3): 345–63. Ballentine, K., and J. Sherman, eds. 2003. The Political Economy of Armed Conflict: Beyond Greed and Grievance. Boulder, CO: Lynne Rienner. Barnes, Catherine. 2009. Renegotiating the Political Settlement in War-to-Peace Transitions. Lon- don: Conciliation Resources. Bell, Christine. 2015. What We Talk about When We Talk about Political Settlements. Edinburgh: Political Settlements Research Programme. Belmont, Katharine, Scott Mainwaring, and Andrew Reynolds. 2002. Introduction. In The Archi- tecture of Democracy: Institutional Design, Conflict Management, and Democracy, ed. Andrew Reynolds, 1–13. Oxford: Oxford University Press. Berdal, M., and D. M. Malone. 2000. Greed and Grievance: Economic Agendas in Civil Wars. Boulder, CO: Lynne Rienner. Bermeo, N. 2002. “The Import of Institutions.” Journal of Democracy 13 (2): 96–110. Binningsbø, Helga Malmin, and Siri Aas Rustad. 2012. “Sharing the Wealth: A Pathway to Peace or a Trail to Nowhere?” Conflict Management and Peace Science 25 (3): 206–23. Boone, C. 2003. Political Topographies of the African State: Territorial Authority and Institutional Choice. Cambridge: Cambridge University Press. Brancati, D. 2006. “Decentralization: Fueling the Fire or Dampening the Flames of Ethnic Con- flict and Secessionism?” International Organization 60 (3): 651–85. Brinkerhoff, D. W. 2011. “State Fragility and Governance: Conflict Mitigation and Subnational Perspectives.” Development Policy Review 29 (2): 131–53. Brown, M. E. 2001. “The Causes and Regional Dimensions of Internal Conflict.” In The Interna- tional Dimensions of Internal Conflict, ed. M. E. Brown, 571–601. Cambridge, MA: MIT Press. Buhaug, H. 2010. “Dude, Where’s My Conflict? LSG, Relative Strength, and the Location of Civil War.” Conflict Management and Peace Science 27 (2): 107–28. Buhaug, H., S. Gates, and P. Lujala. 2009. “Geography, Rebel Capability, and the Duration of Civil Conflict.” Journal of Conflict Resolution 53 (4): 544–69. Buhaug, H., K. S. Gleditsch, H. Holtermann, G. Østby, and A. F. Tollefsen. 2011. “It’s the Local Economy, Stupid! Geographic Wealth Dispersion and Conflict Outbreak Location.” Journal of Conflict Resolution 55 (5): 814–40. 97 Buhaug, H., and J. K. Rød. 2006. “Local Determinants of African Civil Wars, 1970–2001.” Polit- ical Geography 25 (3): 315–35. Castillejo, Clare. 2014. Promoting Inclusion in Political Settlements: A Priority for International Actors? Oslo: Norwegian Peacebuilding Resource Centre. Cederman, L.-E., H. Buhaug, and J. K. Rød. 2009. “Ethno-­ nationalist Dyads and Civil War: A GIS-Based Analysis.” Journal of Conflict Resolution 53 (4): 496–525. Cederman, L.-E., S. Hug, A. Schädel, and J. Wucherpfennig. 2015. “Territorial Autonomy in the Shadow of Conflict: Too Little, Too Late?” American Political Science Review 109 (2): 354–70. Cederman, L.-E., N. B. Weidmann, and K. S. Gleditsch. 2011. “Horizontal Inequalities and Eth- nonationalist Civil War: A Global Comparison.” American Political Science Review 105 (3): 478–95. Cederman, L.-E., A. Wimmer, and B. Min. 2010. “Why Do Ethnic Groups Rebel? New Data and Analysis.” World Politics 62 (1): 87–119. Chapman, T., and P. G. Roeder. 2007. “Partition as a Solution to Wars of Nationalism: The Importance of Institutions.” American Political Science Review 101 (4): 677–91. Cheeseman, Nic. 2011. “The Internal Dynamics of Power-Sharing in Africa.” Democratization 18 (2): 336–65. Colletta, N., and B. Oppenheim. 2017. “Subnational Conflict: Dimensions, Trends, and Options for Prevention.” Background paper for United Nations–World Bank Flagship Study, Path- ways for Peace: Inclusive Approaches to Preventing Violent Conflict, World Bank, Washington, DC. Collier, P. 2001. “Economic Causes of Civil Conflict and Their Implications for Policy.” In Tur- bulent Peace: The Challenges of Managing International Conflict, ed. C. A. Crocker, F. O. Hampson, and P. Aall, 143–62. Washington, DC: United States Institute of Peace Press. Collier, P., and A. Hoeffler. 1998. “On Economic Causes of Civil War.” Oxford Economic Papers 50 (4): 563–73. ———. 2002. “On the Incidence of Civil War in Africa.” Journal of Conflict Resolution 46 (1): 13–28. ———. 2004. “Greed and Grievance in Civil War.” Oxford Economic Papers 56 (4): 563–95. ———. 2005. “Resource Rents, Governance, and Conflict.” Journal of Conflict Resolution 49 (4): 625–33. Collier, P., A. Hoeffler, and D. Rohner. 2009. “Beyond Greed and Grievance: Feasibility and Civil War.” Oxford Economic Papers 61 (1): 1–27. Collier, P., A. Hoeffler, and M. Söderbom. 2004. “On the Duration of Civil War.” Journal of Peace Research 41 (3): 253–73. Cordell, K., and S. Wolff. 2009. Ethnic Conflict: Causes—­ Consequences—Responses. Cambridge: Polity. Crawford, Gordon, and Christof Hartmann, eds. 2008. Decentralisation in Africa: A Pathway out of Poverty and Conflict? Amsterdam: Amsterdam University Press. Csergő, Zsuzsa, Philippe Roseberry, and Stefan Wolff. 2017. “Institutional Outcomes of Territo- rial Contestation: Lessons from Post-Communist Europe, 1989–2012.” Journal of Federalism 47 (4): 491–521. doi: https://doi.org/10.1093/publius/pjx025 Cunningham, D. E. 2006. “Veto Players and Civil War Duration.” American Journal of Political Science 50 (4): 875–92. 98 Subnational Governance and Conflict Deiwiks, C., L.-E. Cederman, and K. S. Gleditsch. 2012. “Inequality and Conflict in Federations.” Journal of Peace Research 49 (2): 289–304. Department for International Development. 2010. Building Peaceful States and Societies. Lon- don: Department for International Development. DeRouen, K., M. J. Ferguson, S. Norton, Y. H. Park, J. Lea, and A. Streat-Bartlett. 2010. “Civil War Peace Agreement Implementation and State Capacity.” Journal of Peace Research 47 (3): 333–46. Di John, Jonathan, and James Putzel. 2009. Political Settlements. Birmingham, UK: Governance and Social Development Resource Centre. Diprose, R. 2009. “Decentralization, Horizontal Inequalities and Conflict Management in Indo- nesia.” Ethnopolitics 8 (1): 107–34. Diprose, R., and U. Ukiwo. 2008. Decentralisation and Conflict Management in Indonesia and Nigeria. Oxford: Centre for Research on Inequality, Human Security and Ethnicity. Doyle, Michael W., and Nicholas Sambanis. 2000. “International Peacebuilding: A Theoretical and Quantitative Analysis.” American Political Science Review 94 (4): 778–801. Ezcurra, R. 2015. “Fiscal Decentralization and Internal Conflict: An Empirical Investigation.” Environment and Planning C: Government and Policy 33 (3): 580–600. Fearon, J. D. 2004. “Why Do Some Civil Wars Last So Much Longer than Others?” Journal of Peace Research 41 (3): 275–301. ———. 2005. “Primary Commodity Exports and Civil War.” Journal of Conflict Resolution 49 (4): 483–507. Fearon, J. D., and D. D. Laitin. 2003. “Ethnicity, Insurgency, and Civil War.” American Political Science Review 97 (1): 75–90. ———. 2008. “Civil War Termination.” Paper presented at the Annual Meeting of the American Political Science Association, Chicago, IL, August 30–September 1, 2007. Felbab-Brown, Vanda, Harold Trinkunas, and Shadi Hamid. 2018. Militants, Criminals, and Warlords: The Challenge of Local Governance in an Age of Disorder. Washington, DC: Brook- ings Institution Press. Filippov, M., P. C. Ordeshook, and O. Shvetsova. 2004. Designing Federalism. Cambridge: Cam- bridge University Press. Forsberg, Erika. 2013. “Evaluating Domino Effects of Granting Territorial Concessions to Sepa- ratist Groups.” International Studies Quarterly 57 (2): 329–40. Gaynor, Niamh. 2016. “Challenges to Decentralisation in the Democratic Republic of the Congo: Beyond the Political Settlement.” Journal of International Development 28 (2): 198– 213. doi:http://doi.wiley.com/10.1002/jid.3046 Goldstone, J. A., R. H. Bates, D. L. Epstein, T. R. Gurr, M. B. Lustik, M. G. Marshall, J. Ulfelder, and M. Woodward. 2010. “A Global Model for Forecasting Political Instability.” American Journal of Political Science 54 (1): 190–208. Gurr, T. R. 1993. “Why Minorities Rebel: A Global Analysis of Communal Mobilization and Conflict Since 1945.” International Political Science Review 14 (2): 161–201. Gurr, T. R., and W. H. Moore. 1997. “Ethnopolitical Rebellion: A Cross-Sectional Analysis of the 1980s with Risk Assessments for the 1990s.” American Journal of Political Science 41 (4): 1079–103. Gutierrez, Eric. 2011. Introducing Political Settlements. London: Christian Aid. Hartzell, C. A. 1999. “Explaining the Stability of Negotiated Settlements to Intrastate Wars.” Journal of Conflict Resolution 43 (1): 3–22. References 99 Hartzell, C., and M. Hoddie. 2003. “Institutionalizing Peace: Power Sharing and Post–Civil War Conflict Management.” American Journal of Political Science 47 (2): 318–32. ———. 2007. Crafting Peace: Power-Sharing Institutions and the Negotiated Settlement of Civil Wars. University Park: Pennsylvania State University Press. Hegre, H., and N. Sambanis. 2006. “Sensitivity Analysis of Empirical Results on Civil War Onset.” Journal of Conflict Resolution 50 (4): 508–35. Hendrix, C. S., and I. Salehyan. 2012. “Climate Change, Rainfall, and Social Conflict in Africa.” Journal of Peace Research 49 (1): 35–50. Hickey, Sam. 2013. Thinking about the Politics of Inclusive Development: Towards a Relational Approach. Manchester, UK: Effective States and Inclusive Development Research Centre. Horowitz, D. L. 1985. Ethnic Groups in Conflict. Berkeley: University of California Press. ———. 2007. “The Many Uses of Federalism.” Drake Law Review 55 (4): 953–66. Humphreys, M. 2005. “Natural Resources, Conflict, and Conflict Resolution.” Journal of Conflict Resolution 49 (4): 508–37. Institute for Economics and Peace. 2018. Global Peace Index 2018: Measuring Peace in a Complex World. Sydney: Institute for Economics and Peace. http://visionofhumanity.org/app/ uploads/2018/06/Global-Peace-Index-2018-2.pdf Internal Displacement Monitoring Centre. 2018. Global Report on Internal Displacement. Geneva: Internal Displacement Monitoring Centre. International Crisis Group. 2018a. “Averting War in Northern Somalia.” Africa Briefing No. 141. https://www.crisisgroup.org/africa/horn-africa/somaliland/141-averting-war-northern- somalia International Crisis Group. 2018b. “Drug Trafficking, Violence and Politics in Northern Mali.” Africa Report No. 267.https://www.crisisgroup.org/africa/sahel/mali/267-narcotrafic-violence-et- politique-au-nord-du-mali Jackson, Paul. 2016. “Local Government and Decentralisation in Post-Conflict Contexts.” Third World Thematics: A TWQ Journal 1 (6): 747–62. Jenne, E. K. 2012. “When Will We Part with Partition Theory? Flawed Premises and Improbable Longevity of the Theory of Ethnic Partition.” Ethnopolitics 11 (3): 255–67. Jenne, E. K., Saideman, S. M., and Lowe, W. 2007. “Separatism as a Bargaining Posture: The Role of Leverage in Minority Radicalization.” Journal of Peace Research 44 (5): 539–58. Jones, Bruce, Molly Elgin-Cossart, and Jane Esberg. 2012. Pathways out of Fragility: The Case for a Research Agenda on Inclusive Political Settlements in Fragile States. New York: Center for International Cooperation. Justino, Patricia. 2017a. “Governance Interventions in Fragile and Conflict-Affected Countries.” IDS Working Paper 496. https://opendocs.ids.ac.uk/opendocs/handle/123456789/13259 ———. 2017b. “Linking Inequality and Political Conflict: The Role of Social Mobilization and Collective Action.” Background paper for the United Nation–World Bank Flagship Study, Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict, World Bank, Washing- ton, DC. Kaufman, S. J. 2001. Modern Hatreds: The Symbolic Politics of Ethnic War. Ithaca, NY: Cornell University Press. Kaufmann, C. 1996. “Possible and Impossible Solutions to Ethnic Civil Wars.” International Security 20 (4): 136–75. Keating, Michael. n.d. “Regionalism and Regionalization.” In Encyclopedia Princetoniensis. https://pesd.princeton.edu/?q=node/264 100 Subnational Governance and Conflict Khan, Mushtaq H. 1995. “State Failure in Weak States: A Critique of New Institutionalist Expla- nations.” In The New Institutional Economics and Third World Development, ed. John Harriss, Janet Hunter, and Colin M. Lewis, 71–86. London: Routledge. ———. 2010. Political Settlements and the Governance of Growth-Enhancing Institutions. Lon- don: School of Oriental and African Studies. Koubi, V., T. Bernauer, A. Kalbhenn, and G. Spilker. 2012. “Climate Variability, Economic Growth, and Civil Conflict.” Journal of Peace Research 49 (1): 113–27. Koubi, V., G. Spilker, T. Böhmelt, and T. Bernauer. 2014. “Do Natural Resources Matter for Inter- state and Intrastate Armed Conflict?” Journal of Peace Research 51 (2): 227–43. Lake, D. A., and D. Rothchild. 1996. “Containing Fear: The Origins and Management of Ethnic Conflict.” International Security 21 (2): 41–75. ———. 1998. “Spreading Fear: The Genesis of Transnational Ethnic Conflict.” In The Interna- tional Spread of Ethnic Conflict: Fear, Diffusion, and Escalation, ed. D. A. Lake, and D. Roth- child, 3–32. Princeton, NJ: Princeton University Press. ———. 2005. “Territorial Decentralization and Civil War Settlements.” In Sustainable Peace: Power and Democracy after Civil Wars, ed. P. G. Roeder, and D. Rothchild, 109–32. Ithaca, NY: Cornell University Press. Lapidoth, Ruth. 1996. Autonomy: Flexible Solutions to Ethnic Conflicts. Washington, DC: United States Institute of Peace Press. Laws, Edward. 2012. Political Settlements, Elite Pacts, and Governments of National Unity. Bir- mingham, UK: Developmental Leadership Programme. Lomborg, B., ed. 2013. How Much Have Global Problems Cost the World? A Scorecard from 1900 to 2050. Cambridge: Cambridge University Press. Lujala, P., N. P. Gleditsch, and E. Gilmore. 2005. “A Diamond Curse? Civil War and a Lootable Resource.” Journal of Conflict Resolution 49 (4): 538–62. Mansell, Wade, and Karen Openshaw. 2013. International Law: A Critical Introduction. Portland, OR: Hart. Marc, Alexandre, Alys Willman, Ghazia Aslam, Michelle Rebosio, and Kanishka Balasuriya. 2012. Societal Dynamics and Fragility: Engaging Societies in Responding to Fragile Situations. Washington, DC: World Bank. Martin, Philip. 2013. “Coming Together: Power-Sharing and Durability of Negotiated Peace Settlements.” Civil Wars 15 (3): 332–58. doi:https://doi.org/10.1080/13698249.2013.842747 Mascagni, Giulia. 2016. “Is Fiscal Decentralisation Delivering on Its Promises? A Review of the Theory and Practice in Developing Countries.” IDS Working Paper 466. https://opendocs. ids.ac.uk/opendocs/handle/123456789/11169 Mattes, M., and B. Savun. 2009. “Fostering Peace after Civil War: Commitment Problems and Agreement Design.” International Studies Quarterly 53 (3): 737–59. McGarry, John. 1998. “‘Orphans of Secession’: National Pluralism in Secessionist Regions and Post-Secession States.” In National Self-Determination and Secession, ed. Margaret Moore, 215–32. Oxford: Oxford University Press. ———. 2007. “Asymmetry in Federations, Federacies and Unitary States.” Ethnopolitics 6 (1): 105–16. McGarry, John, and Brendan O’Leary, eds. 1993. The Politics of Ethnic Conflict Regulation. Lon- don: Routledge. Milanovic, Marko. 2010. “Sejdic and Finci v. Bosnia and Herzegovina.” American Journal of Inter- national Law 104 (4). References 101 Min, E., M. Singh, J. N. Shapiro, and B. Crisman. 2017. “Understanding Risk and Resilience to Violent Conflicts.” Background paper for the United Nations–World Bank Flagship Study, Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict, World Bank, Washing- ton, DC. Mogaka, S. 2017. “Competition for Power in Africa: Inclusive Politics and Its Relation to Violent Conflict.” Background paper for the United Nations–World Bank Flagship Study, Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict, World Bank, Washington, DC. Narang Suri, Shipra. 2016. “Decentralisation and Local Governance in Post-Conflict Contexts: A Practitioner’s Perspective.” Third World Thematics: A TWQ Journal 1 (6): 921–27. https:// www.tandfonline.com/doi/full/10.1080/23802014.2017.1370846 Neudorfer, N. S., U. G. Theuerkauf, and S. Wolff. 2016. “Territorial Self-Governance: An Effec- tive Approach to ­ Territory-Centered Conflict Management?” Paper presented at the Interna- tional Studies Association, 57th Annual Convention, Atlanta, March 16–19. O’Brien, James C. 2010. “The Dayton Constitution of Bosnia and Herzegovina.” In Framing the State in Times of Transition: Case Studies in Constitution-Making, ed. Laurel Miller, 332–49. Washington, DC: United States Institute of Peace Press. OCHA (United Nations Office for the Coordination of Humanitarian Affairs). 2019. “Global Humanitarian Overview 2019.” https://www.unocha.org/sites/unocha/files/GHO2019.pdf O’Leary, B. 2007. “Analysing Partition: Definition, Classification and Explanation.” Political Geography 26: 886–908. Østby, G., R. Nordås, and J. K. Rød. 2009. “Regional Inequalities and Civil Conflict in Sub- Saharan Africa.” International Studies Quarterly 53 (2): 301–24. Parks, T., N. Colletta, and B. Oppenheim. 2013. The Contested Corners of Asia: Subnational Con- flict and International Development Assistance. San Francisco: Asia Foundation. Pearson, F. S., M. O. Lounsbery, S. Walker, and S. Mann. 2006. “Rethinking Models of Civil War Settlement.” International Interactions 32 (2): 109–28. Pettersson, T., and P. Wallensteen. 2015. “Armed Conflicts, 1946–2014.” Journal of Peace Research 52 (4): 536–50. Posen, B. R. 1993. “The Security Dilemma and Ethnic Conflict.” Survival 35 (1): 27–47. Powell, R. 2006. “War as a Commitment Problem.” International Organization 60 (1): 169–203. Rao, Sumedh. 2013. Citizens’ Role in Political Settlements. Birmingham, UK: Governance and Social Development Resource Centre. Rocha Menocal, Alina. 2015. Political Settlements and the Politics of Inclusion. Birmingham, UK: Developmental Leadership Programme. Rode, Martin, Hans Pitlik, and Miguel A´ngel Borrella Mas. 2018. “Does Fiscal Federalism Deter or Spur Secessionist Movements? Empirical Evidence from Europe.” Publius: The Journal of Federalism 48 (2): 161–90. https://academic.oup.com/publius/article/48/2/161/4562682 Roeder, P. G., and Rothchild, D. 2005. Sustainable Peace: Power and Democracy after Civil Wars. Ithaca, NY: Cornell University. Ross, M. L. 2004a. “How Do Natural Resources Influence Civil War? Evidence from Thirteen Cases.” International Organization 58 (1): 35–67. ———. 2004b. “What Do We Know about Natural Resources and Civil War?” Journal of Peace Research 41 (3): 337–56. ———. 2006. “A Closer Look at Oil, Diamonds, and Civil War.” Annual Review of Political Sci- ence 9: 265–300. 102 Subnational Governance and Conflict Ross, Michael W., Päivi Lujala, and Siri Aas Rustad. 2012. “Horizontal Inequality: Decentralizing the Distribution of Natural Resources Revenues, and Peace.” In High-Value Natural Resources and Post-Conflict Peacebuilding, ed. Päivi Lujala and Siri Aas Rustad, 251–59. Abingdon, UK: Earthscan. Rustad, S. A., and H. M. Binningsbø. 2012. “A Price Worth Fighting For? Natural Resources and Conflict Recurrence.” Journal of Peace Research 49 (4): 531–46. Sambanis, Nicholas. 2000. “Partition as a Solution to Ethnic War: An Empirical Critique of the Theoretical Literature.” World Politics 42 (4): 437–83. Sambanis, Nicholas, and Milanovic, Branko. 2014. “Explaining Regional Autonomy Differences in Decentralized Countries.” Comparative Political Studies 47 (13): 1830–55. Sambanis, N., and J. Schulhofer-Wohl. 2009. “What’s in a Line? Is Partition a Solution to Civil War?” International Security 34 (2): 82–118. Schneckener, U. 2002. “Making Power Sharing Work: Lessons from Successes and Failures in Ethnic Conflict Regulation.” Journal of Peace Research 39 (2): 203–28. Schünemann, Julia, and Amanda Lucey. 2015. Success & Failure of Political Settlements: Defining & Measuring Transformation. Edinburgh: Political Settlements Research Programme. Siroky, D. S., and J. Cuffe. 2015. “Lost Autonomy, Nationalism and Separatism.” Comparative Political Studies 48 (1): 3–34. Sisk, T. 1996. Power Sharing and International Mediation in Ethnic Conflicts. Washington, DC: United States Institute of Peace. Stewart, F. 2010. Horizontal Inequalities as a Cause of Conflict: A Review of CRISE Findings. Oxford: Centre for Research on Inequality, Human Security and Ethnicity. Stewart, F., G. Brown, and A. Langer. 2007. Policies towards Horizontal Inequalities. Oxford: Cen- tre for Research on Inequality, Human Security and Ethnicity. Swenden W. 2002. “Asymmetric Federalism and Coalition-Making in Belgium.” Journal of Fed- eralism 32 (3): 67–87. Theuerkauf, U. G. 2010. “Institutional Design and Ethnic Violence: Do Grievances Help to Explain Ethnopolitical Instability?” Civil Wars 12 (1–2): 117–39. ———. 2012. “Ethno-embedded Institutionalism: The Impact of Institutional Repertoires on Ethnic Violence.” PhD thesis, London School of Economics and Political Science. Toft, M. D. 2003. The Geography of Ethnic Violence: Identity, Interests and the Indivisibility of Territory. Princeton, NJ: Princeton University Press. ———. 2010. Securing the Peace: The Durable Settlement of Civil Wars. Princeton, NJ: Princeton University Press. Tranchant, Jean-Pierre. 2016. “Decentralisation, Regional Autonomy and Ethnic Civil Wars: A Dynamic Panel Data Analysis, 1950–2010.” MPRA Paper 72750, University Library of Munich, Germany. UCLG (United Cities and Local Governments). n.d. “Glossary Proposal: UCLG Decentralisation and Local Self-Government Committee.” http://www.uclg-decentralisation.org/sites/default/ files/glossary_uclg_decentralisation_localselfgovernment_committee_eng.pdf UNDP (United Nations Development Programme). 1997. Decentralized Governance Pro- gramme: Strengthening Capacity for People-Centered Development. New York: Management Development and Governance Division, Bureau for Development Policy. ———. 1999. “Decentralization: A Sampling of Definitions.” Working paper, UNDP, New York. http://web.undp.org/evaluation/documents/decentralization_working_report.PDF References 103 ———. 2016. Integrated Framework to Support Local Governance and Local Development. New York: UNDP. UNHCR (United Nations High Commissioner for Refugees). 2018. “Figures at a Glance.” https://www.unhcr.org/news/stories/2018/6/5b222c494/forced-displacement-record- 685-million.html United Nations Document. S/2020/223. Situation in Mali: Report of the Secretary General. 20 March 2020. United Nations Document. S/2019/983. Situation in Mali: Report of the Secretary General. 30 December 2019. United Nations and World Bank. 2018. Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict. Washington, DC: World Bank. Uppsala Conflict Data Program. 2015. Definition of Armed Conflict. Uppsala, Sweden: Uppsala University. USAID (United States Agency for International Development). 2009. Democratic Decentraliza- tion Handbook. Washington, DC: USAID. Wall, Gareth J. 2016. “Decentralisation as a Post-Conflict State-Building Strategy in Northern Ireland, Sri Lanka, Sierra Leone and Rwanda.” Third World Thematics: A TWQ Journal 1 (6): 898–920. https://www.tandfonline.com/doi/full/10.1080/23802014.2016.1369859 Walter, B. F. 1997. “The Critical Barrier to Civil War Settlement.” International Organization 51 (3): 335–64. ———. 1999. “Designing Transitions from Civil War: Demobilization, Democratization, and Commitments to Peace.” International Security 24 (1): 127–55. ———. 2002. Committing to Peace: The Successful Settlement of Civil Wars. Princeton, NJ: Princeton University Press. ———. 2003. “Explaining the Intractability of Territorial Conflict.” International Studies Review 5 (4): 137–53. ———. 2006. “Information, Uncertainty, and the Decision to Secede.” International Organiza- tion 60 (1): 105–35. ———. 2009. “Bargaining Failures and Civil War.” Annual Review of Political Science 12: 243–61. doi:https://doi.org/10.1146/annurev.polisci.10.101405.135301 Weidmann, N. B. 2009. “Geography as Motivation and Opportunity: Group Concentration and Ethnic Conflict.” Journal of Conflict Resolution 53 (4): 526–43. Weinstein, J. M. 2005. “Resources and the Information Problem in Rebel Recruitment.” Journal of Conflict Resolution 49 (4): 598–624. Weller, M. 2008. Escaping the Self-Determination Trap. Leiden, Netherlands: Martinus Nijhoff. Weller, M., and K. Nobbs. 2010. Asymmetric Autonomy and the Settlement of Ethnic Conflicts. Philadelphia: University of Pennsylvania Press. Weller, M., and S. Wolff. 2005. “Recent Trends in Autonomy and State Construction.” In Auton- omy, Self-Governance and Conflict Resolution: Innovative Approaches to Institutional Design in Divided Societies, ed. M. Weller and S. Wolff, 262–70. London: Routledge. Whaites, Alan. 2008. States in Development: Understanding State-Building. London: Department for International Development. Wimmer, A., L.-E. Cederman, and B. Min. 2009. “Ethnic Politics and Armed Conflict: A Config- urational Analysis of a New Global Data Set.” American Sociological Review 74 (2): 316–37. Wimmer, A., and B. Min. 2006. “From Empire to Nation-State: Explaining Wars in the Modern World, 1816–2001.” American Sociological Review 71 (6): 867–97. 104 Subnational Governance and Conflict Wischnath, G., and H. Buhaug. 2014. “Rice or Riots: On Food Production and Conflict Severity across India.” Political Geography 43: 6–15. Wolff, S. 2004. “The Institutional Structure of Regional Consociations in Brussels, Northern Ireland, and South Tyrol.” Nationalism and Ethnic Politics 10 (3): 387–414. ———. 2008. “Complex Power Sharing as Conflict Resolution: South Tyrol in Comparative Perspective.” In Tolerance through Law: Self-Governance and Group Rights in South Tyrol, ed. J. Woelk, F. Palermo, and J. Marko, 329–70. Leiden, Netherlands: Martinus Nijhoff. ———. 2009. “Complex Power-Sharing and the Centrality of Territorial Self-Governance in Contemporary Conflict Settlements.” Ethnopolitics 8 (1): 27–45. ———. 2011a. “Managing Ethnic Conflict: The Merits and Perils of Territorial Accommoda- tion.” Political Studies Review 9 (1): 26–41. ———. 2011b. “Managing Ethno-national Conflict: Towards an Analytical Framework.” Com- monwealth and Comparative Politics 49 (2): 162–95. ———. 2011c. “Post-Conflict State Building: The Debate on Institutional Choice.” Third World Quarterly 32 (10): 1777–1802. doi:https://doi.org/10.1080/01436597.2011.610574 ———. 2013a. “Conflict Management in Divided Societies: The Many Uses of Territorial Self-Governance.” International Journal on Minority and Group Rights 20 (1): 27–50. ———. 2013b. “From Paper to Peace: The Role of Guarantees in Peace Agreements.” World Politics Review, December 17, 2013. Woodman, S., and Y. Ghai. 2013. “Comparative Perspectives on Institutional Frameworks for Autonomy.” In Practising Self-Government: A Comparative Study of Autonomous Regions, ed. Y. Ghai and S. Woodman, 449–86. Cambridge: Cambridge University Press. World Bank. 1999. World Development Report 1999/2000: Entering the 21st Century. Washington, DC: World Bank Group. ———. 2001. “Administrative Decentralization.” http://www1.worldbank.org/publicsector/ decentralization/admin.htm ———. 2011. World Development Report 2011: Conflict, Security, and Development. Washington, DC: World Bank Group. ———. 2014. Intergovernmental Fiscal Relations in Fragile and Conflict Situations. Washington, DC: World Bank Group. ———. 2017. World Development Report 2017: Governance and the Law. Washington, DC: World Bank Group. References 105 About the Authors Stefan Wolff is Professor of International Security at the University of Birmingham, England, UK. He is also an Associate Fellow of the Royal United Services Institute in London and a member of the Advisory Council of the European Centre for Minority Issues. An expert on international conflict management, he has published over 80 journal articles and book chapters, as well as 18 books, including Ethnic Conflict: A Global Perspective (OUP 2007), Conflict Management in Divided Societies: Theories and Practice (Routledge, 2011, with Christalla Yakinthou), The European Union as a Conflict Manager (Routledge, 2012, with Richard G. Whitman), and The Dynamics of Emerging De-facto States (Routledge 2019, with Tetyana Malyarenko). Bridging the gap between academia and policy, he has been engaged in various stages of conflict settlement negotiations, including in Moldova, Iraq, Sudan, and Yemen. Since 2017, he has also worked with the OSCE, especially on confidence building and economic con- nectivity. Wolff holds degrees from the University of Leipzig (BA), the University of Cambridge (M.Phil.) and the LSE (Ph.D.). Simona Ross is a Political Affairs Officer for the United Nations Investigative Team for Accountability of Da’esh/ISIL (UNITAD) in Baghdad, Iraq. From 2016 to 2018, Ross worked with the United Nations Support Mission in Libya (UNSMIL), where she was responsible for issues related to Libya’s political economy and the constitution-making process. Prior to joining the United Nations, Ross worked with the World Bank’s Fra- gility, Conflict and Violence Group and with the World Development Report Team preparing the 2017 WDR on Governance and the Law. Ross completed a research fel- lowship with the Edmond J. Safra Center at the University of Harvard, where her research focused on the influence of institutional corruption on U.S. foreign policy. Ross is completing her graduate studies in International Human Rights Law at the University of Oxford and she is a doctoral candidate at the Department of Politics and International Studies (POLIS) at the University of Cambridge. She also holds a mas- ter’s degree from the SIT Graduate Institute and a bachelor’s degree from the Alpe- Adria University. Asbjorn Wee is the Team Leader for the Recovery, Peacebuilding and Post-conflict Transition Program in the Fragility, Conflict and Violence Department of the World Bank Group in Zimbabwe. He coordinates the WBG’s support to post-conflict needs assessments and supports the WBG country teams with the development of re-­ engagement strategies post-conflict to manage risks associated with conflict and vio- lence. Over the past year he has supported and/or led assessments in such diverse 107 contexts as Colombia, North-East Nigeria, Northern Mali, Myanmar, and Ukraine. From 2007 to 2012, Wee worked at the OECD, where he managed the International Network on Conflict and Fragility and produced guidance for donors on risk taking and how to support post-conflict transitions. From 2003 to 2007, he worked in the WBG’s PREM Network in East Africa in support of the peace processes in Sudan and Somalia and the re-engagement in both countries. He graduated from John Hopkins’ School of Advanced International Studies in 2003. 108 Subnational Governance and Conflict