Identifying Spoiling Behavior and Spoiler Management Strategies: Reconciling Actor-Based and Context-Based Theories

Introduction

The end of the Cold War marked a significant change in the pattern of armed conflicts, as the number of civil wars around the world increased tremendously[1]. Authors and policy makers have consistently reported the “inexorable shift from inter-state to intra-state conflict over time”[2], sometimes conceptualized through the ‘new wars’ thesis[3]. In the 1990s and the first decade of the 21^st century, there has been a significant upsurge in the number of civil wars frequently evolving in correlation with or as a result of an important “breakdown of public authority”[4] and the gradual loss of power by the central state often contested by rebel factions and oppositions. Ramifications of intra-state conflicts for peacemaking are multifaceted for they allow for a greater number of parties involved in wars waged over economic gains or identity grievances rather than ideological objectives[5]. By implicating a myriad of actors, intra-state conflicts and their distinctive characteristics tend to affect the structure of peace processes themselves. Cochrane explains in his book, “the changing pattern in armed conflict has had important effects in terms of how wars are fought, who is affected by them and how they can be brought to an end”[6].

This paper is mainly concerned with the latter part of Cochrane’s observation – ending conflict – because, while the bipolar state of affairs of the Cold War established states as the main interlocutors and actors of peace negotiations, the current dynamics of conflict provides space for non-state actors and private entities such as rebel groups, warlords and militias to participate in the conflict and its resolution. Internal conflicts are consequently more likely to cease through fragile and complex negotiated peace agreements rather than through decisive military victory of one warring party over another. Civil wars are thus fertile grounds for the emergence of actors either reluctant to enter peace processes construed as threats to their very survival while attempting to obstruct negotiations from the outside, or, when involved in a given peace agreement, unwilling to carry out negotiations to a favorable outcome, or their implementation if they perceive such agreements as ultimately detrimental to their objectives. Such behavior is typically referred to as ‘spoiling’ in the literature, and although authors offer slightly conceptually divergent definitions of potential or effective spoilers, they can be broadly identified as parties who seek to hinder peace settlements either during negotiations or their implementation phase, if they deem the process detrimental to their objectives or survival.

Another important attribute of new wars and the resulting peacemaking strategies is that negotiated peace agreements between warring parties have increasingly been facilitated and implemented under the auspices of international actors such as the United Nations (UN), regional organizations and powerful, sometimes neighboring states. These external actors to peace processes are commonly referred to in the literature as “custodians of peace”[7], third parties, and occasionally external peace implementers, essentially tasked with designing and enforcing implementation of peacemaking strategies. The spoiler problem within peace negotiations is thus embedded in the larger theoretical framework of peacemaking rather than peacebuilding, since bringing warring parties at the negotiation table and drafting a peace agreement that will result in sustainable peace are efforts that usually arise during the first stages of a peace process. It is during negotiation rounds and the following short-term implementation efforts that spoiling behavior is most likely to occur. By way of contrast, peacebuilding typically develops at later stages of the peace process since the “term presupposes that there is a peace to build upon”[8] and aims at longer-term strategies including reconciliation, transitional justice, institutional and economic reconstruction and so on.

Briefly recounting the recent changing patterns in conflicts and the subsequent effects on peacemaking is key to understanding the conditions in which the custodians, and internal parties to the peace evolve, and why they need to pay heed to potential spoiling behavior that could wreck the process altogether. Apprehending the spoilers’ motives, capacities to spoil and the context within which they evolve is a critical step for peace brokers to take upon before designing and engaging in ‘spoiler management’ strategies. Such endeavor has proved uneasy as the very term ‘spoiler’ as well as its ‘labeling’ function not only prove problematic theoretically, but also incur practical implications in the process of choosing the adequate spoiler management tactic. The subsequent “spoiler management” strategies might prove ill-designed, unrealistic or simply too detached from the realities of the conflict. In order to assess the significance of such challenges, this paper will first study the remaining challenges present in the current literature on the spoiler problem in peacemaking. In fact, a close look at the conflict resolution scholarship reveals two competing models aimed at identifying and managing spoilers: the first one is personality-oriented and seek to define spoilers by their internal organization and characteristics of their leadership; the second explores the impact of context, potentially shaped by third-parties, on the emergence of spoilers in peace processes. The second part of this article will thus focus on showing that although each model has its separate departing point, their associated spoiler management strategies are not mutually exclusive and could prove all the more effective if efficiently combined by policy makers.

Literature Review: the Challenges of Defining and Identifying Spoilers

Conceptualizing spoilers in peace processes: narrow versus broad definition

Drawing from his experience as “an observer in Angola in 1992”[9], Stedman in his pioneering work on spoilers[10] observed that the previous literature on robust guarantees and credible commitment in peace implementation tended to focus on confidence-building strategies and – wrongly – assumed that parties joining peace negotiations would do so in good faith, and were ‘genuinely interested in peace, but none can trust the others sufficiently to keep promises”[11]. Stedman provided peacemakers with a major contribution to the existing literature on peacemaking and spoiling behavior for he was the first scholar to insist on the fact that all parties are not acting in good faith. In effect, some actors deliberately enter peace talks for what Newman and Richmond refer to as ‘devious objectives”[12], that is, for tactical reasons such as buying time to regroup, or avoid international pressure and criticism, and pursue their specific, pre-negotiation objectives. Stedman thus pointed out that “confidence-building was absolutely ineffective if one or more sides decided to return to war and block implementation of a peace agreement”[13]. In his first article, Spoiler problem in peace processes, he attempted to define and label spoilers as “leaders and parties who believe that peace emerging from negotiations threatens their power, worldview, and interests, and use violence to undermine attempts to achieve it”[14].

Such definition can be read as narrow because it excludes an important aspect of spoilers and spoiling behavior: it does not say anything about actors who do not use direct violence[15] in peace processes and choose to stall negotiations using other means. Cochrane for example, argues for a different terminology and defines spoilers as “resisters of negotiated peace settlements”. This way, one can “examine the range of motivations and strategies used by those who seek to resist political settlement”[16]. Cochrane’s endeavor, in line with Newman and Richmond, is to include in his definition the possible use of both violent and peaceful means not to merely shatter the peace process, but sometimes to shape it. Moreover, his critiques contend that Stedman’s definition assumes that spoilers’ behavior is based on a “simplified binary discourse” of “those who are ‘for’ and ‘against’ conflict settlement” [17]. In fact, evidence shows that peace processes are not that simple and contain major grey areas in which actors’ behaviors towards the peace process change according to “environmental variables”[18]. Stedman’s definition thus tends to ignore potential or “latent spoilers” who might emerge any time along a given peace process if they are provided with the opportunity to spoil[19]. Moreover, labeling spoilers in a narrow way tends to rest on normative judgments[20], hence affecting the way peace brokers think of spoiler management strategies, and can result in misreading a spoiler’s intentions. In response to such shortcomings, peacemaking scholars have sought to flesh out their argument on a broader definition for spoilers. For instance, Richmond and Newman conceive ‘spoiling behavior’, rather than labeling reluctant parties as ‘spoilers’, as a flexible, fluctuating process that might stem from “any actors who are opposed to peaceful settlement for whatever reason from within or (usually) outside the peace process and who use violence or other means to disrupt the process (…)”[21].

The broad versus narrow definition dichotomy directly leads to the conception of two main models. Stedman’s actor-based model draws on his narrower definition, thus focusing on spoilers’ “position in the peace process, number of spoilers, type based on intent, and whether the locus of spoiling behaviour lies with the leader or the followers (…). Of crucial importance is the motivation and intent of the spoilers”[22]. Stedman hence describes the spoiler problem based on actors and their pre-negotiation motives. His critiques consequently produced a different, context-oriented model which posits that all actors are potential spoilers, who change tactics and will decide to spoil the peace process or not, while using violence or voice, depending on the ‘opportunity structure’ in which they evolve. Although Stedman repeatedly criticized the existing literature for providing undifferentiated analyses of armed conflicts, his actor-oriented model might not be broad and developed enough to account for all the various elements conducive to spoiling behavior. Indeed, even though actors’ motives and internal agency of armed groups are important, the context in which they evolve will play a critical role in understanding conflicts and planning for adequate peacemaking strategies. Thus, the novelty of the context-based model is that it allows for better-differentiated analyses of armed conflicts dynamics and the prospective for peace. Since the model develops in light of the environment specific to every internal conflict, it helps to understand the fact that actors to the conflict and their objectives can be fluctuating depending on the general context, thus permitting peace brokers to identify latent spoilers ex ante. The following part discusses in further details these two approaches to spoiling in peace processes.

Spoiling behavior: the actor-based model

As briefly outlined above Stedman developed his argument around specific items relating to spoilers in peace processes and their capacity to spoil. Stedman insists that his study of spoiling is “a first step toward understanding the spoiler process” and recognizes that “some of the findings will need to be reconsidered and revised”[23]. In his quest to depart from undifferentiated studies of armed conflicts and their resolution, he contends that spoilers are not a homogeneous group as they differ by their goals and commitments; an effort aimed at designing better strategies applicable by the custodians of peace “for keeping peace on track” in specific conflict zones[24]. He thus identifies three types of spoilers – limited, greedy or total – that vary depending on two dimensions: their goals (limited or total) and the level of commitment dedicated to achieving them (low or high).[25] Limited spoilers have limited objectives such as specific grievances, which can be integrated into the peace settlement. However, limited goals do not imply low levels of commitment. Total spoilers function on an “all-or-nothing” basis and usually display high levels of commitment and violence as peacemaking might directly threaten their survival. Greedy spoilers lay between limited and total spoilers and have limited goals that will expand or contract depending on cost and benefit calculations and the peace agreement capacity to absorb their demands. Authors who criticized Stedman often make the case that his typology rests on the underlying hypothesis that spoilers whether limited, greedy or total, will remain as such during a given peace process, hence assuming that contextual patterns will not matter much in altering spoiling behavior[26]. Zahar notes that even “a given peace agreement can fluctuate wildly” and offers the example of the PLO’s attitude towards the peace process in Israel between 1993 and 2001, in which the Yasser Arafat continuously changed behavior thus obscuring his intentions and strategies and blurring the lines between total or greedy categories.

Stedman’s typology does not only include a first step to identifying types of spoilers; he goes further as to determine other dimensions of the spoiler problem that confront custodians of peace. In fact, similarly to Zahar’s viewpoint[27], he recognizes the importance of the “position of the spoiler” in the peace process. On the one hand, an inside spoiler has agreed to enter the peace negotiations and “yet fails to fulfill his key obligations to the agreement”[28]. On the other hand, outside spoilers are denied participation to the peace process – or exclude themselves – and seek to destroy it from the outside by using violence. A third dimension of the spoiler problem rests in the fact that miscalculations by the custodians of peace regarding the number of spoilers can “inadvertently strengthen another”[29]. Here, he uses the Arusha peace accords (Rwanda, 1993) as an illustration, where the UN, by coercing President Habyarimana into complying with his obligations, directly buoyed up the outsider CDR[30] to spoil the peace process from the outside. The UN failed to recognize the presence of a strong outside spoiler by dealing with only one of them, where they were in fact, two[31]. Finally, Stedman describes the locus of the spoiler as being an important factor in understanding which strategy might work best. This aspect of spoilers’ behavior puts the emphasis on leadership and followers and directly relates to the possible shifts in type of spoilers. It is certainly important for custodians of peace to avoid overshadowing the potential changes in leadership and the associated conduct of its followers[32].

Most importantly, Stedman identifies three ‘spoiler management strategies’ that apply depending on the types of spoilers, namely coercion, inducement or socialization strategies[33]. In short, “whereas limited or greedy spoilers can be addressed through inducement or socialization, total spoilers can only be managed with a coercive strategy”[34]. Such strategies will be discussed further in the ‘analysis’ section of this article, but it is critical to recall that they must be supported by third parties and designed according to the identified type of spoilers. Overall, it is crucial that custodians use the right combination of sticks and carrots to either integrate or fight outside spoilers, and to accommodate or contain insiders. Although Stedman shows that such strategies should not be adopted in a rigid fashion and might be prone to change, this is not so because of the effect of evolving context surrounding spoilers, but rather due to their intrinsic characteristics and types. The first step to identifying adequate spoiler management strategies is thus a correct diagnosis of spoiler types[35] by the custodians of peace. According to the author, poor diagnosis will inexorably lead to failed spoiler management. 

Spoiling behavior: the context-based model

Adopting a more flexible definition implies that, contrary to Stedman’s more rigid model of spoiler types; spoiling behavior is subject to change depending on structural factors rather than actors’ personality or personal preferences. Along with Richmond and Newman’s assertion that spoiler issues are contextual and depend on ‘environmental variables’[36] Greenhill and Major develop a ‘capabilities-based model’[37] by adopting a reverse standpoint to Stedman’s analysis: spoilers will not emerge as such solely based on their own personalities and preferences, but rather as a result of the opportunity structure and the distribution of power. In this view, all spoilers can become equally greedy depending on the conflict environment in which they operate; they are not borne, but made. Actually, “the type of spoiler does not determine the type of outcomes that are possible; instead, the kinds of outcomes that are possible determine the type of spoiler that may emerge at any given time”[38].

Therefore, the custodians of peace’s primary task is to apprehend the distribution of power among the competing factions shaping the conflict dynamics rather than determining the relevant type of spoilers they face.  To Greenhill and Major, departing from the opportunity structure – to select the right spoiler management strategy – rather than the spoilers’ motives and intent is better suited than Stedman’s suggestion because the “degree of spoiler behavior (…) is largely a reflection of the probability that a given actor assigns to the likelihood that he can unilaterally alter the situation on the ground and the level of risk he is willing to assume to do so, rather than a fundamental difference in type”[39]. As a result, it is essential that peace brokers account for spoilers’ “devious objectives”[40] that is, analyzing the vantage point of disputants, which includes the possibility for would-be spoilers to gauge the potential opportunities to spoil the process if the benefits of going back to the battlefield exceeds those of the peace dividends. This “inside-out” approach also sanctions the need to account for spoiling behavior that translates into non-violent action and appears to be legitimate “according to an alternative rationality”[41] differing from that of the custodians of peace.  In fact, the parties’ pre-negotiation goals must be properly understood for third-parties to realize that entering a peace process might be for the warring factions a “continuation of the conflict by other means”[42] including strategies of stalling and delaying the process not necessarily involving overt violence.

The first step to identifying the right spoiler management strategy is thus to acknowledge the nature of the conflict, the variables affecting the opportunity structure and the subsequent change in the distribution of power that will draw spoilers to enter cost and benefit evaluations of spoiling. In the same vein, Zahar’s argument draws from Stedman’s assertion that the location of spoilers is critical to spoiler management; that is if they are inside or outside the peace process[43]. Indeed, considering such distinction will help custodians to understand the reasons why spoilers emerge and how they assess the costs ad benefits of spoiling. For instance, insiders might decide to comply with the settlement provisions because both the costs of returning to war and losing the peace dividends are too high. However, if they see the peace agreement as a threat to their objectives, they might try to delay, obstruct or stall the process, and refuse to comply during the implementation phase. On the contrary, in the viewpoint of outsiders, the costs of entering the peace process will be too high if it jeopardizes their very survival, and the loss of their territorial and economic gains for instance, especially as peace agreements hold provisions for strict demobilization and disarmament measures. Although her argument focuses on the outside versus inside spoiler dichotomy, her analysis shows, like Greenhill and Major as well as Richmond and Newman, that costs and benefit calculations of spoilers do not depend on their types, but rather on the external context and outcomes, including shifts in the opportunity structure and distribution of power. The context-based model although designed as a criticism of Stedman’s typology does not reject his contribution because his analysis of elites in peace processes remains an important complement to his critiques’ broader stand. In fact, a context that evolves with third party intervention or changes in leadership for instance will arguably shape the warring parties’ incentives to spoil, but cannot be construed separately from their pre-negotiation objectives, intentions and personality of their leaders.

Analysis: Reconciling Actors-Based and Context-Based Models

Before entering the details of Stedman’s spoilers management strategies and his critiques’ propositions to remedy to potential spoiling behavior, it is important to recall why specific context of civil wars play a role in engaging with armed groups and potential spoilers and what are the ramifications of a continuously unstable environment for peace processes.

Context matters: the characteristics of ‘New wars’, third parties and associated challenges

Richmond and Newman develop the ‘new war’ thesis as a critical component to understanding spoiling behavior. Indeed, heterogeneous non-state armed groups sustain a critical role in the progress of intra-state conflicts, and can either be potential spoilers or effectively engaged in spoiling behavior during peace negotiations. They operate in the context of “new wars”, or “contemporary conflicts”, that feature specific characteristics such as state failure, massive population displacements, the targeting of unarmed civilians, and shadow economies, thus violating international humanitarian law and creating highly composite conflicts that are difficult to stop and prevent[44]. Conflict resolution therefore becomes an intricate task in situations where certain actors have acquired a vested interest in the continuation of war as they “find incentives to the continuation of violence, public disorder and the political economy of war”[45]. Characteristics of new wars thus help us understand the context in which spoilers choose to either engage in spoiling behavior or decide to comply with a given peace agreement. Such contextual considerations come to buttress Greenhill, Richmond and Zahar’s argument that opportunity structures and environmental variables are critical elements in creating spoilers.

Here come in considerations of external actors may they be custodians of peace or spoilers’ patrons. Third parties involved in the peace process and its short-term implementation might affect the opportunity structure, and thus the emergence of spoilers. Greenhill and Major[46] as well as Zahar[47] point at the “credible commitment problem” that can arise when international custodians do not show strong capability and commitment to ensure implementation by all parties and retaliate in cases of non compliance. By taking the Khmer Rouge example in Cambodia, Stedman observes that depending on UN peacekeepers ability to engage in protecting civilians and punishing spoiling behavior, disputants may be emboldened to resume violence or forced to give up arms and observe compliance[48]. Moreover, Zahar discusses the ‘security dilemma’ whereby spoilers show reluctance to demobilize and disarm, and demonstrates the crucial role custodians hold in ensuring compliance of armed groups to their obligations[49].

In practice however, third party intervention in the peace process is not always positive. Most authors point at other contextual factors such as the presence of hostile neighbors supporting and fueling spoiling behavior before and during peace processes, can be determining in the creation of spoiling behavior[50]. This is important because it shifts the focus of custodians of peace on non-state entities to states acting as spoilers, directly or not, for political and tactical reasons. A case in point is Ethiopia’s constant efforts to mold peace settlements in Somalia, influence the structure of the country’s new transitional governments and sometimes destroy the process when considered contrary to Ethiopia’s own interests. This reinforces the hypothesis that the context-specific model mentioned above might be an appropriate practical tool and an important refinement of Stedman’s conceptual framework for designing spoiler management strategies.

Confronting and reconciling actor-based and context-based spoiler management strategies

The problem with the current literature on spoiling behavior rests in the fact that Stedman’s typology, although recognized as an essential tool in managing spoilers, has been largely criticized on grounds of practical concerns: his spoilers typology does not allow predictive measures since it prevents the identification of spoilers ex ante. Yet, criticism has largely glossed over the innovation he brought through his specific spoiler management strategies. In fact, a lot has been done to refine his three spoiler types as a departing point for spoilers management strategies, but little has been done to assess the inducement, socialization and coercion strategies laid down in his article, essentially aimed at uncovering the specificities of given internal armed conflicts, their actors’ objectives and their reactions to the unfolding peace process. Contrary to his opponents’ general statements on context-related strategies such as ‘sticks and carrots’ and deterrence through third-parties’ credible commitments, Stedman sought to go beyond this type of generalization in order to conceptualize specific means to deal with spoilers thus avoiding undifferentiated analyses of conflict-resolution strategies.

As briefly evoked above, there are three possible tactics to deal with spoilers in peace processes: inducement, socialization and coercion. According to Stedman, inducement consists in accommodating the warring parties’ grievances. They can derive from fear of being deprived of security in the case of demobilization and disarmament for instance, fairness of the peace negotiations outcomes and justice, that is “demand of legitimation or recognition of their position”[51]. Such grievances will obviously be construed as legitimate from the vantage point of the parties to the peace agreement, it thus the third party responsibility to enter the dangerous task of assessing how legitimate they are and the ways by which they can be addressed, rejected or accommodated. Even though inducement is considered to be a “default mode” because it does not necessarily takes into account the relevant spoilers type, Stedman’s article shows that it is best applied in the case of limited spoilers. Socialization – best suited for greedy and limited spoiler type – aims at persuading the parties of the validity of “a set of norms” based on a material and an intellectual component. The latter includes democratic values such as fair elections and the “protection of human rights” whereas the former consists of a balance between sticks and carrots used to “reward and punish the spoiler”[52]. According to Stedman’s article, an example of successful inducement is the case of Mozambique, where RENAMO, “a South-African trained and assisted guerilla movement”[53], and the government of Mozambique entered a peace negotiations under the auspices of the UN in 1992 followed by elections. Departing from the premises that RENAMO and his leader fell under the limited spoiler category, the UN, the US as well as regional powers – south Africa and Zimbabwe – followed a strategy of mixed inducement and socialization, alternatively agitating the right amount of sticks and carrots. Mozambique is the only instance of socialization (used concurrently with inducement tactics) laid down in Stedman’s article, proving how difficult such a strategy can be in a conflict environment not necessarily conducive to respecting human rights and understanding the democratic principles of free and fair elections. Finally, Stedman identifies two types of coercive strategies – applicable to total spoilers – aimed at decreasing the spoiler’s capacity to spoil: the ‘departing train’ and ‘withdrawal’ strategies[54]. The ‘withdrawal’ strategy, used by the UN in Rwanda in 1993 was a complete failure as the CDR, construed as a total spoiler, took the opportunity to foster its ideological stands and carried out the genocide a few months later. The ‘departing train’ strategy was carried out in Cambodia against the Khmer Rouges – also considered total spoilers – following the 1991 Paris peace accords. As the Khmers Rouges continued to spread terror and carry out violent attacks against the population and governmental forces to destroy the peace process and the upcoming elections, the UN efficiently reorganized its UNTAC peacekeeping forces and endeavored to contain the Khmers Rouges in order to “protect the electoral process”[55], by ensuring that elections took place, with or without them.

At the other end of the spectrum, Greenhill and Major’s attempt to assess the probability of spoiler emergence is based on a model of “predicted level of greed for a given distribution of power”[56]. In short, the distribution of power among competing factions of the peace process is flexible; it can shift rapidly depending on four factors that can catalyze spoiling behavior. The first and second factors are directly linked to the ‘commitment problems’. If the international community supporting the peace process does not commit to strong levels of “independence, will, means, and mandate to countercyclically intervene in the ongoing implementation of the accords”[57], spoiling behavior might emerge as a party interpret the lack of ‘credible commitment’ by the custodians of peace as an opportunity to spoil the process. Moreover, credible commitments guarantee that “the balance established during negotiations will be maintained during the implementation phase of the accords, convincing combatants to demobilize their armies and surrender conquered territory”[58]. The third and fourth factor directly take root in the potentially changing domestic political context, prompting parties, both confronted to a mutually hurting stalemate, to enter negotiations. In any case, “none of these four   (assumptions) is intention driven or personality specific”[59], but they greatly contribute to the emergence or avoidance of spoiling behavior since the “actors’ willingness to negotiate coincide with changes in the distribution of power and with the demonstrated readiness of the custodians of peace to change the opportunity structure on the ground”[60]. Contrary to Stedman’s specific management strategies applicable to spoilers, Greenhill and Major’s context-based model for spoiler management draw from the more general concepts of traditional deterrence, mutually hurting stalemates[61] and the credible commitment problem[62] applicable in the broader context of civil wars.

Recounting such conceptualized strategies is important, not merely because Stedman and Greenhill connects them respectively to spoiler types and context-dependent spoiling behaviors but because a proper analysis of their study cases show that the strategies he describes are compatible with and complementary to his opponents’ criticism. Both Greenhill and Major’s article on spoilers illustrate their points by comparing actor and context-based models through Stedman’s major examples – Cambodia, Angola and Mozambique –in an effort to confront each other’s standpoints. Studying the relevance of each comparative analysis is beyond the scope of this paper; however the case of Angola is worth discussing because it shows not only that both analyses and strategies are not mutually exclusive, but also, that a combination of elements from both studies might lead to effective spoilers management. Stedman’s analysis of the peace process focuses on the leaders’ personality and the subsequent category of spoiler they fall under, while Greenhill and Major insist on the broader context, in which third party action, by shaping the opportunity structure, ultimately induce the spoilers to change their line of conduct. In his analysis of the characteristics of the conflict resolution strategies implemented in Angola, Stedman equates the failure of the third parties’ inducement strategy to the incorrect assessment of the types of spoilers they were facing during the peace process. UNITA leader, Jonas Savimbi and President of Angola, dos Santos engaged in the 1991 Bicesse peace accord, which provided for immediate cease-fire, demobilization and disarmament, the subsequent integration of MPLA (government forces) and UNITA forces into a single military, and the organization of free and fair multiparty elections[63]. Unfortunately, Savimbi rejected the 1992 elections outcome as incumbent dos Santos took the lead. The UN, and most specifically the US failed to properly understand that Savimbi “characterized his conflict in all-or-nothing terms” and sought to exert power over the entire Angolan territory rather than accept a power-sharing agreement in times where the balance of power seemed to lean in his favor. According to Stedman, Angola and the other two cases of inducement, Mozambique and Cambodia, show that this tactic only works with limited spoilers. If spoilers fall under the category of greedy rather than limited, inducement will have the unfortunate reverse effect to embolden spoiling behavior[64]. By giving greedy spoilers what they want, the benefits of demanding more concessions and strategic gains by stalling negotiations will outweigh the benefits of respecting the terms of a peace agreement that might not serve their interests as well. Yet, Savimbi was seen by the US administration in Washington as a limited spoiler while he was in fact greedy; this poor diagnosis and consequently selecting the inadequate spoiler management strategies led to a swift return to violence after the elections took place. Greenhill and Major on the other hand contend that Savimbi’s behavior only altered once the opportunity structure had changed and led him to accept a return to the negotiating table. They note that although they concur with Stedman’s diagnosis that place Savimbi under the ‘greedy’ label, the outcome of the Bicesse peace process did not depend on his “personality and ambitions” but rather on the “combination of a weak UN, access to external funding and supplies, and a disintegrating MPLA (which) created an opportunity structure that Savimbi – once UNITA lost the elections – was bound to exploit”[65]. However, this only comes to reinforce Stedman’s viewpoint on spoiler management strategies since inducement did not work in reality. Moreover, these authors all refer to the shift in the balance of power between UNITA and MPLA at a critical moment of the peace process. Indeed, just like Greenhill and Major, Stedman asserts that “the transition period had upset the balance of power between the two militaries” while he further recognizes the critical role played by Savimbi’s closed links with Washington through his powerful lobbying strategies, the important revenues he gained from his exploitation of diamond mining sites and the lack of resolve emanating from the UN to insist on form a contingency plan in case Savimbi lost the elections and failed to recognize their outcome as legitimate[66]. The major aspect differentiating the two analyses on Angola lies in the fact that Greenhill and Major believe that had the third party to the Bicesse peace settlement engaged in better strategies showing strong commitment to monitor and implement the peace agreement, they might have succeeded in alleviating Savimbi’s recalcitrant behavior. This is an important aspect for the purpose of this paper because it shows how third parties can alter the environmental variables in a way that will provide the warring parties with disincentives to spoil the process by raising the costs of returning to war. This demonstrates that the capabilities-based model “allows one to predict ex ante if and when spoilers might emerge or retreat”[67]. Their conclusion nonetheless, cannot be reasonably disconnected from Stedman’s plea for custodians of peace to carry out a thorough investigation in order to reveal the character, original motivations and personal background of a given spoiler group’s leadership. In the Angolan example, Stedman rightly points at the fact that US officials in Angola, who experienced the situation in Angola first hand, realized that inducement would not work in light of “Savimbi’s personality and ambitions”[68] and his inclination to see the world through the lens of “racism, paranoia and megalomania (that) led him to believe that the MPLA had stolen the elections from him and that he had the right to rule all of Angola”[69]. Major and Greenhill explain the outcomes of the peace processes in Mozambique and Cambodia on similar grounds as they contend that commitment to deterrence strategies only worked out so long as third parties successfully increased the costs for potential spoilers to derail the peace process and insist that grounding spoiler management tactics on spoiler types fails to explain the success of failure of peace processes.

These illustrations of spoiler management tactics in building peace and generating compliance to peace processes show that accounting for the specific context is critical in choosing the relevant strategies to avert spoiling behavior. They also lead to further questions associated to other relevant context-specific aspects of conflict resolution strategies. For instance, in order to shift the debate from spoilers’ leadership characteristics to context-oriented strategies, Zahar points out inconsistencies in Stedman’s typology. She remarks that greedy and limited spoilers can sometimes “share the same characteristics” while the “same spoiler types often span two cells of the typology”[70]. In other words, Stedman’s model only permits to distinctively separate total spoilers from the rest, a category that is usually difficult to apply in the field, as pointed out earlier in this paper. This leads to the conclusion that spoilers are in reality all potentially greedy depending on context-variable costs and benefits calculations. However, Zahar recognizes that “security dilemmas and commitment problems offer an undifferentiated analysis of civil wars”[71], and fail to properly account for specific spoiler issues. In light of this, if third-parties are to be actively involved in peace processes, they need not pay attention to specific types of spoilers, but rather “stay attuned to the context in which they are operating” because, in line with Greenhill and Major’s argument, they have the opportunity to affect the “opportunity structure available to would-be spoilers” [72]. Furthermore, a closer look at Zahar’s article permit to connect the dots between context and actors’ devious objectives. Interestingly, her analysis of inside and outside spoilers’ incentives to spoil is based on an intricate calculation of costs and benefits, trade offs and the possible loss of the peace dividends for would-be spoilers. Evaluating the opportunity structure surrounding them will also take into account the degree to which latent spoilers are attached to their pre-negotiation objectives, or devious objectives. Although the concept of devious objectives developed by Richmond share certain characteristics of the context-based model, it allows for an important connection between spoilers’ personalities closely linked to their pre-negotiation objectives. If such goals can be identified early on in the peace process, spoiling attitudes might be averted especially if third-parties understand that such objectives do not necessarily disappear with bargaining and negotiations endorsed by third-parties[73]. Finally, both Zahar and Richmond, among others, reveal the downside of third parties acting as molders of the opportunity structure, because external parties do not always act effectively to prevent spoiling behavior, or not at all. As Stedman pointed out custodians of peace such as UN peacekeepers often suffer from “organizational blinders” and might be slow both to assess the situation on the field and respond to immediate threats. Although this might be disputable, UN missions are in theory neutral components of peace implementation, contrary to neighboring states’ intervention such as Syria’s meddling into Lebanon’s internal affairs. According to Zahar, “these actors are not only often partial to one faction, they are also less philosophically opposed to the use of force in implementation”, which can have disastrous effects on the peace process as Ethiopia interference in Somalia peace processes sadly showed. In a later work, Stedman fully recognized the importance of handling and avoiding violence during peace processes and consequently concurred with his critiques in terms of the credible commitment problems. In fact, he takes a slightly different stand as he insists that one of the main components of successful peace implementation is the sufficient amount of resources and attention that third parties dedicate to certain cases if their national interests are directly affected by the conflict[74]. Similarly to Zahar and others, he further asserts: “international role in implementation should be fine-tuned to the difficulty of the implementation environment” [75].

A contemporary example of spoiling behavior and credible commitment issues is the ongoing violence in Syria. Syrian President al-Assad although supposedly committed to implement the six point peace plan negotiated by Koffi Annan and supported by a UN observer mission deployed last April[76], both the government and the rebel armed forces have systematically breached the cease fire while Damascus has refused to withdraw heavy weaponry from populations center. Although the situation on the ground remains difficult to assess, it is clear that, like Stedman explains in the last section of his article, lack of consensus and coordination within the UN Security Council, the Arab League and the major neighboring countries has led the Syrian government to engage in deceptive tactics. His initial willingness to accept negotiations and implement to the six-point plan seems like a strategic move to gain time, delay international criticism, blur his true motivations and objectives and continue using violence against the Syrian population. This is an interesting case both in terms of spoiler types and potential management strategies. If applied correctly, Stedman’s typology would contend that President Assad as well as the rebel Free Syrian Army are both total spoilers as they hold an all-or-nothing view on their conflicts – respectively regime change and continued control over the entire Syrian territory – and display high levels of commitments since they are both ready to fight each other using overt violence. Another interesting element of the Syrian case is that in designing the peace talks, neither Anna, the UN Permanent 5 nor the Arab League deliberately attempted to include the Free Syrian Army, thus sending the message that they did not consider the latter as a valid interlocutor, nor expected the rebels to engage in spoiling behavior. Peace brokers thus face a situation in which both inside and outside spoilers fight to wreck the peace process, most likely construed as illegitimate or discredited. However, the fact that Assad has accepted to enter negotiations shows that, if properly compelled by the third parties’ forceful action, he might have been inclined to respect the cease fire and comply with his obligations. In line with the context-based model, this makes him as greedy as any other potential spoiler since the opportunity structure, here characterized by poor international coordination and the absence of agreement among the main peace brokers, allows him to ignore the peace plan. It seems like we know too little about the rebel army to understand their primary motives and personality, but in all cases, it seems like inducement might be the least appropriate strategy to deal with the Syrian case. This is the final connection with Stedman’s pioneering work as he insisted: “a common denominator among the successful cases of spoiler management is unity and coordination among external parties in defining the problem, establishing legitimacy for the strategy and applying the strategy”[77]. A coercive strategy may then have to be carefully planned by custodians, by taking into account Assad’s motivations and possibly his personality. In addition, the strategy should properly consider and reinforce the potentially critical role of the Arab League and the UN in providing leverage to take action in the face of such manifest violations of the peace agreement. Early on, Stedman noted, “the biggest potential liability (yet source of possible leverage) in managing a spoiler are member states that are patrons of the spoiler”. Discordant opinions among the P5 in the midst of the Syrian crisis is a telling example.

Concluding Remarks

Overall the main criticism against Stedman’s viewpoint is that his typology does not allow for an identification of potential spoilers ex ante. Stedman believes custodians of peace should first indentify spoiler types before engaging in spoiler management strategies while others defend the hypothesis that disputants engage in spoiling behavior depending on outcomes and opportunities. This ‘capabilities-based model” permits custodians to shape the opportunity structure on the ground in order to prevent and anticipate spoiler emergence. Stedman’s work should not be dismissed however, because his analysis of spoilers’ personal characteristics can still be of great practical use in dealing with leaders and understanding their goals and motives. In addition, his inside/outside spoiler distinction allows custodians to anticipate spoiling behavior depending on disputants’ location in the peace process.

Another important point is that, contrary to Stedman, his critiques do not comprehensively answer the question of what type of spoiler management should be deployed on the ground. Despite Greenhill and Major’s emphasis on classic deterrence to keep spoilers at bay through credible commitment to ensure compliance[78], the literature does not always shed light on detailed, potentially effective spoiler management strategies. It remains nonetheless that third parties should engage in a correct diagnosis and thorough analysis of both the context of a given internal conflict and its main actors’ motives and objectives if adequate spoiler management strategies are to be efficiently carried out. This article thus showed that actor-based and context-based analyses of the spoiler problems in peace processes should not be construed as two distinctive, divergent models but rather as complementary and mutually reinforcing. If ingredients from both models can be consciously and systematically used in spoiler management strategies, custodians of peace might strengthen their capabilities to ensure better compliance to peace settlements and their implementation.

Further Reading on E-International Relations

• A Rules-Based System? Compliance and Obligation in International Law
• UN Intervention: Help or Hindrance in the Democratic Republic of Congo?
• Composing Compositeness: Examining EU Defining Actor Aspect in Russia Relations
• The Implications of Stabilisation Logic in UN Peacekeeping: The Context of MINUSMA
• The Future Challenges Facing Europe as a Global Actor
• EU Migration Policy: The EU as a Questionable Actor and a Realist Power