A Post-Cold War Package for Peace

Kaldor (2007a) argues that the end of the Cold War saw the increased importance of what she calls ‘new wars’. These are intra-state wars centred on identity politics, they involve non-state actors, are financed through transnational networks, and civilians are the primary targets. While I am unconvinced of the newness of such wars (see Berdal, 2003), we clearly did see an upsurge after the end of the Cold War and the international responses also changed. Firstly, there was a greater international willingness to intervene and an expansion of the remit of peacekeepers. Peacekeeping was no longer just about separating armies; the 1990s and 2000s saw the creation of international missions that aimed to rebuild war-torn states and in a number of cases took over the running of these states until stability was ensured (or this was at least the idea) (see e.g. Bellamy & Williams, 2010; Gutteri & Piombo, 2007). Secondly, the solution proposed by these international interveners came to be described as the ‘liberal peace’. The emphasis was on introducing political and economic reforms and thereby creating (multi-ethnic) democracy and market economy, which was seen as the key to stability. However, this approach has been heavily criticized by the critical peace-building literature for reflecting global power inequalities and for being insensitive to local culture, identity and history (see e.g. Mac Ginty & Richmond, 2013).

This debate has generally focused on the post-settlement period and has largely ignored the agreement that preceded it. Peace agreements tend to be seen simply as ‘contextual or permissive conditions for post-conflict activity’ (Selby, 2013, p. 64). However, other authors have shown greater interest in peace agreements and argue that the tool kit available to mediators has expanded since the end of the Cold War, and more creative and dynamic approaches are now available. Weller (2008, p. 156) argues that we have escaped the ‘self-determination trap’ and that the range of possible solutions to separatist conflicts has been significantly enhanced, while Bell (2008) identifies a rapidly evolving new Law of Peace, which includes both a ‘new law of hybrid self-determination’ and a ‘new law of transitional justice’. But the question is to what extent these solutions are widespread or remain the exception: are we predominantly dealing with creative, liberal agreements? Moreover, we still lack knowledge of the way in which the different elements of an agreement interact and the conditions under which they are likely to succeed.

In the existing conflict resolution literature we find a rich debate on the relative merit of different institutional designs, including autonomy and federalism, and forms of power-sharing (see e.g. Lapidoth, 1996; Reilly, 2001; Roeder & Rothchild, 2005). But these mechanisms tend to be treated separately not analysed as a package (Cordell & Wolff, 2009, p. 138) and the focus is primarily on their impact on long-term stability rather than how they affect the prospect for getting the conflict parties to agree to a settlement in the first place. These questions are addressed in some quantitative studies but, although very valuable, these studies do not provide a detailed analysis of the content of peace agreements. On the issue of territorial autonomy, we are for example often left with a dichotomy: decentralization or no decentralization,¹ but this does not fully capture the differences between these agreements and the importance of, sometimes rather subtle, details.

When it comes to peace talks, the tendency in the literature is to focus on the process and the context rather than the content. Zartman’s (2001a) concept of a ‘ripe moment’ for peace has proved hugely influential, and the emphasis when explaining the success or failure of peace talks tends to be on the relative military strength of the two sides, the availability of external support, or the existence of spoilers (see Stedman, 1997). In empirical cases of failed peace talks, the typical argument is similarly that what is needed is a change in the conflict context. The agreement did not fail because of its content; no amount of institutional creativity will resolve the situation. The content of peace agreements is therefore under-analysed – at least if we look at the agreement as a package – and the crucial interaction between content, context and process is rarely subjected to in-depth scrutiny.

This book fills a gap in the literature by providing a comprehensive and systematic analysis of the content of peace agreements and the ways in which this content interacts with the conflict context,² and the process of the peace talks. In other words, the content of the peace agreement matters: peace is not simply a question of ‘ripeness’, nor can crucial decisions be left to the post-settlement phase, even in case of significant third party involvement. The agreement, the institutions it creates and the actors it empowers, significantly constrains what is possible at a later stage. This book stresses that peace agreements must be viewed as a whole: the different elements interact and this affects both the extent to which different arrangements are acceptable to the conflict parties and the sustainability of the agreement. A full analysis of the effects of different institutional designs, moreover, necessitates an understanding of intra-communal dynamics, including the political contestation and governance capacity found within the separatist movement. This is not usually afforded much attention in existing literature on peace agreements.

The following chapters find that autonomy – usually territoriallyand ethnically-defined – is an overwhelming trend in peace agreements signed in separatist conflicts, while human rights are surprisingly marginal. The illiberal nature of many settlements can in part be explained by the ‘core deal’, and its prioritization of collective rights, but also owes something to the narrowness of the typical peace process and an assumption that such ‘soft concerns’ can be left for later. The agreements demonstrate some ingenuity when it comes to trans-border dimensions, but the dominant, binary, conception of sovereignty and statehood still poses significant constraints and the examples of ‘simple autonomy’ are more numerous. The analysis warns against ‘destructive ambiguity’, exclusive agreements and a lack of sub-state capacity. It suggests that a greater emphasis on intra-communal dynamics and political contestation within conflict parties could help temper the tendency of peace agreements to reflect unduly the narrow interests and perspectives of the negotiating elites. A broadening of the peace process and of the resulting institutional design could result in a more legitimate and therefore more sustainable peace.

Bridging the Gap: Peace Agreements and Case Studies

The book combines a systematic analysis of peace agreements signed in self-determination conflicts after the end of the Cold War with more in-depth case studies and examples. It thereby endeavours to bridge the current gap between quantitative and qualitative studies of peace agreements. The peace agreements have been selected according to two criteria. Firstly, they attempted to find a solution to intra-state conflicts that involved a confrontation between a sovereign independent state and a rebel movement that sought self-rule or outright independence. The self-determination movement is in most cases ethno-nationalist, but there are exceptions, such as the Sudanese People’s Liberation Movement/Army in South Sudan. Secondly, they are comprehensive agreements that were signed by major parties in the conflict and aimed to address the underlying causes of the conflict.

The focus is limited to self-determination, or separatist, conflicts in order to make an in-depth analysis manageable and since these conflicts are more directly comparable than other types of intra-state conflicts. It has moreover been argued that although separatist wars are often ‘unending’ and violence tends to reoccur, they more commonly end in negotiated settlements than other types of civil wars (Heraclides, 1997). Finally, by only focusing on self-determination conflicts that all involve a territorial element, it is possible to analyse the interaction between all four types of mechanisms – relating to territory, security, power and justice. Some of the findings will however also have implications for other types of intra-state conflicts, especially since solutions proposed in these conflicts sometimes also include a territorial element, such as a federal model for Syria (Reuters, 2016).

The second criterion, a comprehensive peace agreement, is commonly used in peace agreement datasets.³ My interpretation is however less strict than what is used by, for example, the Peace Accords Matrix. Firstly it allows for some substantive issues to be left for later, as this can be an important negotiation technique that for example proved successful in the case of the Belfast Agreement for Northern Ireland. It also allows for a peace agreement to take the form of an agreed change to the constitution, such as in the case of Ukraine–Crimea. Secondly, all potential ‘spoilers’ are not necessarily included in the agreement: spoilers do not always succeed in undermining an agreement and I am specifically interested in how spoiler activity, and its effectiveness, is influenced by the content of the agreed settlement. But I do require both sides to have accepted the agreement – even if under considerable pressure. The Comprehensive Proposal for the Kosovo Status Settlement, also known as Ahtisaari’s Plan, is therefore not included, even though it was later implemented by the Kosovo Government. Determining what is, and what is not, a comprehensive agreement is of course not always straightforward. I chose for example to include the very brief Khasavyurt Accord for Chechnya, although it contains very few institutional mechanisms and is deliberately vague on the issue of status. Including this agreement helps me examine possible solutions at the extreme end of the spectrum – in cases where the separatist movement has achieved de facto independence – and the problem of postponing crucial issues.

Finally, unlike most datasets on peace agreements, I also include settlements signed in non-armed conflicts, including Serbia & Montenegro, Moldova–Gagauzia, Ukraine–Crimea and Russia–Tatarstan.⁴ These conflicts all posed significant challenges to the state and there was a clear potential for violence, as evidenced by later developments or by other separatist conflicts facing the state. Including these agreements allows me to examine the effect of violence, or its absence, on possible solutions.

This results in a list of twenty agreements signed between 1990 and 2010 (see table 0.1). In addition to an in-depth analysis of these main settlements, I also examine any subsequent agreements which added details to the original agreement or included new actors. Such as the Interim Agreement on the West Bank and the Gaza Strip (Oslo II, 1995), in the case of Israel–Palestine, and the Comprehensive Agreement on the Bangsamoro (2014), in the case of Mindanao–Philippines. Later developments, in particular implementation difficulties, also form a core part of the analysis.

This list of agreements tells us two things about separatist conflicts. First of all it points to the surge in these conflicts in the early 1990s, which explains why most of the settlements were concluded in this decade. Secondly, although the list covers a range of different conflict contexts and a wide geographic domain – including cases from Europe, Africa, Asia and Australasia – the lack of cases from Latin America is noticeable. This does not reflect an absence of intra-state conflicts or settlements on this continent, but it does reflect an absence of separatist conflicts. I will briefly analyse agreements signed in non-territorial conflicts, including in Latin America, when examining the effect of territorial provisions on issues relating to security, power and human rights, but the focus of this book is on territorial conflicts. The list is intended to be complete, but it is possible that I have left out agreements that others would characterize as comprehensive or that I am missing certain conflicts with a separatist element.⁵ Intra-state conflicts are fluid: conflicts that were initially about who should govern can acquire a territorial dimension and come to involve the continued existence of the state. However, even if a couple of cases are missing, the list should be comprehensive enough to identify key trends and common problems, and analyse the interaction between different parts of the agreements and the effects of different conflict contexts.

Tabulation is used to provide an overview of these agreements, but this is combined with detailed analysis of their content. This makes the analysis different from existing datasets. For example, I examine the specific form of territorial autonomy (extent, along what lines, how guaranteed?) and the interaction between the different mechanisms. In addition to the analysis of the content of these agreements, a number of failed peace proposals are also scrutinized. These add to the analysis of the effects of the conflict context: what are the limits to different institutional approaches, when do they provide a possible solution? Comprehensive proposals put forward in six cases of self-determination conflicts are examined: Georgia–Abkhazia, Cyprus, Serbia–Kosovo, Azerbaijan–Nagorno Karabakh, Sri Lanka, Morocco–Western Sahara.

The analysis of the content of peace agreements and proposals is combined with more in-depth analysis of specific cases. The purpose of this is to analyse more closely the interaction of the different elements of an agreement and between the peace agreement, the peace process and the conflict context. It will, in other words, allow for a fuller analysis of what explains the failure and success of peace agreements. The case studies draw on extensive research that I have conducted over the years in the Balkans and the Caucasus, but additional examples from other regions will also be included. These examples draw on secondary sources and I have, as far as possible, triangulated any information obtained from these. This did present a challenge with some of the conflicts that have only received limited media and scholarly attention, and a few errors and misapprehensions can therefore likely be found by area specialists. However, I base none of my conclusions on one case only, let alone just one source, so this should not undermine the main thrust of my argument.

Evaluating Success and Failure

Any attempt to evaluate peace agreements should ideally make use of fairly well-defined success criteria. All the twenty agreements are successful in the short-term in the sense that it was possible to get major parties to agree to them. An additional success criterion commonly used is that they have lasted for five or ten years. Both of these will be referred to in the analysis. However, it is not always clear what it means for an agreement to have lasted. It may be fairly clear that an agreement has collapsed if full-scale war breaks out anew, but what if significant provisions were not implemented, or if only some factions continued their violent campaign? Hartzell and Hoddie (2007, p. 137) describe the 1996 peace agreement for Mindanao as a success, since hostilities between the Philippine Government and the Moro National Liberation Front (MNLF) did not reignite. I would however characterize it as a failure since the strongest armed faction, the Moro Islamic Liberation Front (MILF), was not included, several provisions were left unimplemented, and war broke out again in 2000 (see e.g. International Crisis Group, 2013a). The 1997 Chittagong Hill Tracts Peace Accords is even more difficult to classify; most observers agree that significant parts of the agreement have never been implemented, but although tensions remain and violence occasionally flares up, the separatist conflict has not reignited (see e.g. Fortna, 2008). Another problem is deciding at what point an agreement fails. It is for example hard to find anyone who would argue that the Oslo Accords for Israel–Palestine were successful, but at what point did they fail: was it when key provisions were left unimplemented, when Rabin was assassinated and Netanyahu was elected new Israeli prime minister, when Hamas launched a series of suicide bombings in Israel, or was it not until the collapse of the Camp David talks in 2000 and the outbreak of the Second Intifada? Did it therefore pass the five-year threshold, or not? It is consequently difficult to decide on clear-cut criteria for success, and for failure. This is made even more difficult by the post-settlement violence that characterizes many cases but does not necessarily spell the end of an agreement.

The benefit of not doing a large-N study is that it allows for a more nuanced analysis. The specific threshold is not crucial, nor are binary understandings of success and failure. I am interested in institutional provisions that lead to tension, and sometimes violence, and also in the quality of the peace that results; not simply in whether an agreement makes it past an arbitrary threshold. An agreement may survive despite significant tensions, but these tensions still hold important lessons for the effects of different provisions. Moreover, some agreements that did make it past the ten-year mark eventually broke down, which suggests that they did not succeed in addressing the underlying causes of the conflict. The kind of regime that results from the agreement is another important concern: have they ensured ‘good enough governance’, which some authors argue we should settle for, or have they managed to achieve a ‘legitimate peace’ which is supported both horizontally and vertically (Themner & Ohlson, 2014)? This matters for the longer-term sustainability of the agreement and it matters for the conditions faced by the civilian population. Finally, it should be noted that secession will not be considered a failure if the agreement allows for the break-up of the state and the separation does not result in the outbreak of widespread violence. The purpose of a peace agreement is to ensure peace, not the continued existence of the state.

A Word on Terminology

In the analysis I tend to refer to either ‘self-determination conflict’ or ‘separatist conflict’ and I use ‘separatist forces’ and ‘rebel forces’ interchangeably. The term separatist does come with negative con-notations; it is used by central governments to suggest a lack of legitimacy and even illegality. It is therefore unsurprising that the self-proclaimed authorities (another value-laden term) in Donetsk in Eastern Ukraine declared: ‘we are not separatists’ (Walker, 2014) and prefer to be known as rebels. I therefore agonized over whether or not to use this term and use it without any value judgement implied. Linguistically, ‘separatist’ often works better as an adjective than ‘self-determination’ and ‘rebel’ can suggest that what is at stake in the conflict is the nature of the government. And these forces are of course separatist in the sense that they are trying to separate from the central government and achieve at least extensive self-rule.

Plan of the Book

The first part of the book examines the content of the agreements. It analyses how they have addressed contentious issues relating to territory, security, power and justice; discusses advantages and disadvantages of these approaches; examines how these mechanisms are affected by the conflict context, such as the degree of violence experienced; and finally analyses how the different mechanisms interact. Chapter 1 focuses on territory and finds that autonomy is clearly the dominant solution, even if this is sometimes an interim solution and may include attempts to fudge sovereignty. Such attempts are constrained however by dominant conceptions of sovereignty. Chapter 2 examines security provisions found in the agreements; it argues that rather than a centralized monopoly of force, which is typically recommended in the literature, we find complex compromises and frequent creation of autonomous coercive forces. Legitimacy as well as effectiveness matters and this helps address the commitment problem often emphasized in the literature, and helps explain why a robust third party guarantee is not always needed. Chapter 3 focuses on power and finds that (political) power-sharing is much less frequent than we might expect; instead what we see is territorial and military power-sharing. It discusses the limitations of the dominant power-sharing approaches and suggests that a hybrid model may provide a way forward, although this does necessitate third party involvement. Chapter 4 examines if the ‘core deal’ found in these agreements, with its emphasis on territorial autonomy and rights for the dominant groups, leaves any space for justice and rights for individuals and non-dominant groups. It finds that human rights are much less emphasized than what the ‘liberal peace’ literature would lead us to believe; these agreements may be written in the language of human rights but are lacking in substance. Chapter 5 sums up this analysis and asks if we can identify a post-Cold War blueprint for peace. It argues that there are commonalities but the agreements also respond to the conflict context and we are therefore at most dealing with blueprints, in plural. The chapter emphasizes the importance of viewing agreements as a package, when examining the effect of specific provisions; the sometimes destructive impact of ambiguity; the importance of intra-communal dynamics and sub-state capacity; and the difficulty of moving beyond the initial agreement.

The second part of the book focuses on the interaction between the conflict context, the peace process and the content of the peace agreement. Chapter 6 addresses the internal dynamics of the conflict: the balance of power between the two sides and the intra-communal balance between hardliners and moderates. It stresses that both ripeness and the existence of (effective) spoilers are, at least in part, a product of the peace process and the content of the agreement, and proposes an enriched concept of ripeness, which takes into account intra-communal dynamics and is focused on the sustainability of a negotiated settlement. Chapter 7 examines the external side of the equation and discusses the effect of third party involvement. It emphasizes the positive contribution that third parties can make, even if the involvement is not ‘robust’. But there are clear limitations and one of the main constraints is the content of the settlement. The conclusion summarizes the findings and discusses policy implications.