‘De Mistura ideas to reality are like windmills to Don Quixote’: A UN mediator in the Syrian conflict

ABSTRACT

This article assesses the agency of UN mediators through a single case-study of the UN mediator in Syria – Staffan de Mistura. Drawing on and contributing to an emerging research agenda, we argue that UN mediators do have significant room of manoeuvre and thus the ability to conceive, negotiate, and execute key mediation policy-decisions. Our analysis bridges the literature on levels of analysis in International Relations (IR) and Foreign Policy Analysis (FPA). We developed a method that uses process-tracing to identify distinct mediation initiatives. We then pinpoint the mediator’s input on the conceptual and operational components of these main pillars of the mediation process. We question the prevailing perspective wherein the individual mediator has become conspicuous by their absence as the research gaze has been geared towards contextual constraints to mediation, the organization employing the mediator, and developing generic guides for good practice in ‘successful’ mediation. We put the study of the mediator’s agency firmly back into the comparative study of mediation, this way strengthening the validity of arguments that mediators’ personalities, skills and their individual characteristics shape their mediation initiatives just as they point up to considerable personal responsibility for which they need to be better held to account.

KEYWORDS:

• Conflict mediation
• civil war
• Syria
• United Nations
• international organizations

Despite their prestige and expert status, the degree of influence United Nations (UN) mediators have as decision-makers remains underexplored. This article assesses mediators’ individual agency by way of a single in-depth case-study of the degree of influence the UN mediator in Syria – Staffan de Mistura – enjoyed in conceiving, negotiating, and executing mediation policies advanced from his appointment in July 2014 until his resignation in October 2018. Our analysis bridges the literature on levels of analysis in International Relations (IR) and Foreign Policy Analysis (FPA) by allowing for an in-depth case study exposing the degree of individual agency decision-makers have within international organizations. Contributing to an emerging research agenda, we argue that UN mediators do have significant room of manoeuvre and thus the ability to conceive, negotiate, and execute key mediation policy-decisions.

Pinpointing the agency of mediators exposes the influence and leeway mediators, operating within the decision-making structure of the UN and the dynamic context of conflict, have over a mediation process. In brief, it determines the input of the individual mediator into the process of mediation. While our focus may ultimately benefit from the larger structure-agency debates in IR theory and FPA, here we more modestly hope to contribute to the literature on decision-making in mediation and conflict studies. In doing so, we push back against classical framings of mediation that tend to obscure the degree of agency and responsibility mediators have over the policies that shape the mediation processes they oversee. In addition, our study showcases a systematic way to empirically establish the agency of decision-makers in foreign policy that is relevant for theoretical approaches to IR and FPA that attribute significance to first-level analysis as their main premise.

Our effort to bring the individual mediator firmly back into the study of mediation may help strengthen comparative studies of mediation and pave the way for developing potential typologies of what now appear as mediators’ idiosyncrasies – or as one scholar put it, their tendency to just ‘wing it’.¹ We argue that a better understanding of variations in mediation can be gained by grounding the study of mediators’ decision-making at three levels of policy making: conception, negotiation, and execution even while acknowledging that these often overlap. Accordingly, we draw on and contribute to growing insights within FPA and IR theory in assessing how agency allows for and affects mediation policies and decisions. We complement approaches within the study of mediation focused on mediators’ skills and styles that mostly assume but do not demonstrate individual agency as their main premise. At a policy-prescriptive level, our work underscores the need for methodologies to determine accountability of mediators and contribute to developing them.

To identify De Mistura’s input into each of the UN’s mediation policies in Syria and to understand the agency of mediators as decision-makers more generally, this article deploys tools in FPA and process-tracing, and draws on dozens of semi-structured interviews with former UN officials and aids working on Syria, international diplomats, Syrian negotiators, members of Syrian civil society, human rights advocates, and other relevant stakeholders, held between 2017 and 2019. Many interviewees requested to be anonymous. Where possible, these interviews are corroborated with official documents, UN resolutions, reports, speeches, and public statements by the mediator, as well as news sources.

Following a discussion of the literature on agency and first-level analysis in the study of mediation and international policy-making, we make the case for studying the agency of UN mediators. We then turn to the De Mistura’s mediation efforts by identifying the key initiatives during his term in office to then in detail explore his individual footprint in the mediation process and assess alternative perspectives that would discount his agency. We aim for a better understanding of the agency mediators have as decision-makers and making a case for our suggested method by which to assess such agency and influence. That, we argue will contribute to the study of mediation and feed into the debate about mediators’ accountability.

First-level analysis and the study of mediation

Extant literature rarely demonstrates a full appreciation of the significance of individuals and their agency in mediation despite some scholars having hinted at the need to more prominently incorporate such first-level analysis.² A prevailing approach views mediation as predominantly determined by contextual factors. Most notably, approaches to mediation focused on ‘mutually hurting stalemates’ or ‘ripeness’ of conflicts typically look at contextual variables to explain the successful resolution of conflicts, and in their absence its failure.³ Rarely are such contextually driven frameworks marked by an appreciation of the mediator’s ability to manoeuvre and to contribute to a resolution of the respective conflicts they are supposed to mediate. Alternatively, mediators are often regarded as unitary rational actors from a perspective wherein individuals and their strategic perceptions are considered to be largely insignificant to the policy-making process.⁴ The latter would thus understand the mediator – in our case, in the Syrian conflict – to be the UN; consequently attributing little or no importance to the individual occupying the post of mediator. Most notably, quantitative studies have largely framed mediators as unitary rational actors as relevant international actors like the UN are here commonly treated as the unit itself, leaving little room for analysis of the decision-making of the individual mediating on behalf of the organization.⁵ To the extent that the literature on mediation does consider personal attributes and idiosyncrasies of the meditator, the latter informed characterizations of mediators’ required skills and competence.⁶ In such studies the gaze is directed at the mediator, but in their often generic treatment incognizant of his individuality and mostly without demonstrating the varying degrees of leeway mediators effectively enjoy. Related perspectives on ‘strategies’ and ‘styles’ of mediation more often recognize mediators’ individuality but they too rarely demonstrate that the latter indeed matters by way of an analysis of whether and how mediators can and do exert agency in the first place to effectuate their strategies.⁷ This seems further to be the case due to a preoccupation with how to measure a ‘success’ or ‘failure’ of mediation.⁸ This debate certainly is relevant to scholars interested in explaining whether or not mediation brought about an end to conflict, but less so to those who first try to understand the main sources that determined the policies pursued by the mediator irrespective of or prior to assessing their ultimate impact on the conflict.

To be fair, even if not explicitly dedicated to first-level analysis of mediators, case studies of mediation efforts in specific conflicts, including the case of Syria,⁹ have analysed the behaviour of individual mediators and not just the organizations they represent. Yet most of these works are by such mediators themselves or their direct aides in autobiographical accounts of their own mediation efforts.¹⁰ Worthwhile as a source, these works are inherently biased, and not necessarily geared towards pinpointing the room for manoeuvre or agency of the individual mediator let alone his or her policymaking responsibility. Other works that focus on the individual mediator, including recent studies on Syria, are primarily policy prescriptive in nature.¹¹ Like the literature on general mediation skills mentioned earlier, they present valuable insights but mostly do not demonstrate the centrality of a mediator’s individuality as a unit of analysis and principal agent by which such prescriptions could gain relevance or become feasible. In short, there still is a paucity of studies with a consistent focus on mediators’ agency and based on a robust method to assess their leeway, influence, and significance.

From this perspective, the study of mediation should take heed of discussions within IR and FPA on levels of analysis and broader attempts to within their fields of study acknowledge and understand individual agency. Famously, both Waltz and Singer introduced three levels of analysis by which to examine and explain phenomena in the field of IR – the individual, the national, and the systemic.¹² While Waltz and Singer recognize that most IR analysis draws on all the different levels of analysis, they noted that ultimately all studies prioritize one over the others. This way a focus on individuals has long been recognized as valid and even necessary. However, more than two decades ago Byman and Pollack bemoaned what they described as the overwhelming neglect in the study of IR of first level analysis. They challenged theoretical arguments levelled against first-level analysis, and provided criteria to bolster studies aimed at clarifying the impact of individuals on international relations.¹³ Ever since greater appreciation of the relevance of individuals in policymaking has marked a nascent literature in the study of decision-making in conflict settings as well as more established work in FPA and IR theory.¹⁴ One important perspective explored the complex interplay between agents and structures, finding inspiration in political psychology which draws particular attention to the significance of leaders’ personalities and traits.¹⁵ This prompted work drawing on process tracing as an effective tool to assess the significance of leadership traits in decision-making.¹⁶ In doing so, decision-makers are implicitly recognized as pivotal units of analysis, and their role within broader social and political theory is underscored. For instance, Hermann’s research on Leadership Trait Analysis (LTA), following earlier work by Carlsnaes, positions personality and leadership style within the intricate dynamics of the social and political environment wherein leaders operate and decisions are made.¹⁷ More recently, LTA delved into the impact of ‘leaders’ in contexts where ‘structures’ may influence decision-makers to varying degrees.¹⁸ This research draws on process tracing and case studies to identify ‘occasions for decision’,¹⁹ drawing our focus to instances when policymakers are involved in and the authors of a sequence of decisions.

Collectively, the focus on personality and the diverse traits of leaders, their interaction with institutional and social structures, and the strategic use of process tracing can fortify an analytical framework geared towards bringing individual agency back into the study of mediation. Our approach aims to move beyond the implied recognition of agency and make explicit the centrality of decision-makers as units of analysis to explain how decisions are made in UN mediation efforts. While the study of personalities and traits in this respect offers important insight and opportunities, a deliberate emphasis on agency enhances our ability to rigorously assess the degree of responsibility wielded by a decision-maker over a particular course of action. This nuanced distinction holds particular significance for future research and policy formulation directed towards scrutinizing the accountability of mediators.

We may begin to appreciate the importance of mediators’ individual agency by recalling Byman and Pollack’s study on the seminal role of decision-makers in shaping policy-making more generally. Central to their thesis is that certain environmental conditions accentuate the power of the decision-maker, thus suggesting explanatory significance of the first level of analysis. Specifically, they claim that ‘[i]ndividual personalities take on added significance when power is concentrated in the hands of a leader, when institutions are in conflict, or in times of great change’.²⁰ First, in environments where leaders wield significant decision-making power, they are freer to directly shape – if not determine – the main conceptual and organizational aspects of policymaking, such as its agenda, its framing and process, and its finalization. Second, institutional, systemic, and organizational forces, all conceivably reduce the power of the decision-maker. However, when such forces interact with one another or are loosely defined, they often weaken one another to increase the margin of manoeuvre of the policymaker. Decision-makers can thus take advantage of contexts to advance their own interests and agendas. Third, in more dynamic and fast-paced environments, the role of decision-makers becomes more pronounced. Individuals are faster in responding to crises than bureaucracies, and will often take to the foreground if expediency is required. These observations broadly resonate with other studies within FPA and IR prioritizing the first level of analysis and revealing environmental factors that augment individual decision-makers’ significance.²¹ There thus seems to be some agreement that the conditions of conflict or crisis – that these are chaotic, stressful, and dynamic – accentuate the power and thus the explanatory significance of individual decision-makers. As mediators appear to operate in exactly such settings, there is thus a strong case to focus on the individual decision-maker inhabiting this role.

There is yet a method to be developed that allows analysts to study the agency of a mediator in the mediation process meaningfully and feasibly. We take the actual decisions made during the policy-making process, in our case that of the mediation efforts, as the outcomes under scrutiny. The key initiatives during the period the mediator served are the outputs that we will seek to identify and explain. In order to accurately grasp of these decisions, process tracing is used to identify and confirm the most significant decisions to be studied. This allows for a critical discussion of the explaining variables for these outputs and provides a first step towards clarifying the extent of influence the decision-maker had on determining the decision, as opposed to contextual variables derived from the environment or factors emanating from within the UN hierarchy. In doing so, the mediator’s footprint becomes apparent to the extent it will serve a thorough analysis of their agency. To be clear, by agency, we are referring to the influence and leeway of the mediator as a decision-maker over the mediation process.

We draw on Jervis’ observation that to determine the relevance of the agency of decision-makers we must illustrate the conditions that point up to this relevance and rule out counterarguments that challenge the significance of the decision-maker.²² In other words, we must identify what ‘agency’ looks like – or how the individuality of the mediator affects the decisions that made up the mediation process. In order to trace the mediator’s agency, we examine the influence of the mediator on the formulation of each major mediation policy at a conceptual and operational level. The outputs under study this way resemble but, by including an operational level, go beyond Hermann’s ‘occasions for decisions’ that respectively entail the stipulation of broad policy directions, goals and objectives and choosing particular action and commitment of resources.²³ Thus, by studying the extent to which the mediator shaped the conceptual foundations of each mediation policy and the subsequent negotiation and implementation of that policy, we capture the extent, scope, and degree of influence of the mediator at the multiple levels of policy formation. To further elucidate on the mediator’s agency, we assess proposed or conceivable alternative perspectives that, implicitly or explicitly, suggest that mediators act like unitary rational actors, or that structural and contextual variables take analytical precedence. Specifically, we need to rule out that mediators are constrained by path dependency, as in being stuck by the legacy established by their predecessors, at the expense of their individual ability to act, innovate, or to divert from preconceived scripts. We also need to demonstrate that mediators’ decisions do not fully result from the ways in which they socialized – neither in the process by which they are selected, nor upon taking on their new positions. To do so, we demonstrate the variance between De Mistura and his predecessors, Kofi Annan and Lakhdar Brahimi, in their approaches. All three mediators have distinct backgrounds: Annan, a former UN Secretary General with considerable diplomatic capital, Brahimi one of the organization’s veteran mediators with extensive experience working in the region, and De Mistura with more of a background in the UN’s humanitarian operations. To rule out the socialization of the mediator as predeterminant in his decision-making, we highlight the absence of group-think within the UN’s mediation team and draw attention to important differences between the UN mediator and his advisors and technical members of his team. In so doing, we explore the agency of De Mistura and demonstrate how alternative explanations discounting the mediator’s agency are to be viewed as less plausible.

De Mistura’s mediation efforts

De Mistura kicked off his mediation with a regional tour, during which he described himself as being in ‘listening mode’.²⁴ He stated that he was pursuing a limited ‘action plan’.²⁵ On 30 October 2014 a glimpse emerged of what this action plan consisted of as he proposed his first initiative to the UNSC, the ‘Aleppo Freeze’ – a UN mediated ‘freeze zone’ that intended to deescalate the violence in Aleppo, open the way for humanitarian access, and build trust among the Syrian parties.²⁶ Local ceasefires had been largely negotiated without the UN and against the backdrop of the UN failing to prevent and break sieges, primarily imposed by the Syrian regime. The Aleppo Freeze coincided with a government campaign to retake the city as regime troops surrounded rebel held areas in Aleppo and, by 2015, were remounting an offensive against Eastern Aleppo, with a focus on cutting off its supply lines. For months De Mistura continued to pursue consultations to advance the Freeze in Aleppo, but to no avail.

When after five months the ‘Aleppo Freeze’ failed to gain traction, UNSG Ban Ki Moon asked De Mistura to resuscitate the political process. Toward this end, De Mistura advanced a new initiative – the ‘Geneva Consultations’, a series of loose, preparatory meetings with various Syrian stakeholders. The Geneva Consultations opened the door for Syrian civil society groups to participate in the mediation process in a way they had not been previously invited to.²⁷ De Mistura continued to build on his openings with civil society made during the Geneva Consultations by establishing ‘advisory boards’, including a Syrian Women Advisory Board and a Civil Society Support Room.

Before the Intra-Syrian Talks commenced in early 2016, De Mistura participated in two high-level multilateral meetings, co-sponsored by the U.S. and Russia, and held in Vienna between October and November 2015 that included 20 foreign ministers but no Syrians. The first meeting concluded with the ‘Vienna Communiqué’, calling for a diplomatic solution and singling out IS as a shared threat to be defeated.²⁸ It called on the UN to mediate the next round of talks between the Syrian parties.

In November 2015, taking on the name of International Syria Support Group (ISSG), the ISSG agreed to the ‘Vienna Statements’, stipulating a peace plan to resolve the conflict in Syria. While the fate of Assad in Syria in a transitionary process remained unresolved, the parties agreed to supporting a nationwide ceasefire in the country (notably exempting military operations against IS and Jabhat al-Nusra, the al-Qa’ida-affiliate in Syria), following through with the implementation of a Syrian-led political transition as agreed in the Annan-brokered Geneva Communiqué of June 2012, and arranging formal negotiations between the Syrian parties on 1 January 2016. The agreement entrusted De Mistura to oversee these negotiations. On 18 December the UNSC adopted Resolution 2254 endorsing the communiqués of Vienna and Geneva and asking De Mistura to lead the next formal round of negotiations between the Syrian parties.

Instead of convening a central peace conference on Syria as was done in Geneva I (June, 30, 2012) and II (between January and February 2014), De Mistura held nine rounds of largely indirect talks between the Syrian parties through him – the ‘Intra-Syrian talks’.

The first round of intra-Syrian Talks began on 29 January 2016 in Geneva and was to focus on ‘the establishment of a broad ceasefire, the delivery of humanitarian aid, as well as stopping the threat posed by Da’esh [IS]’.²⁹ Yet the attempt faltered as Russian air strikes and regime advances significantly worsened humanitarian conditions in Latakia and Aleppo, prompting De Mistura to ‘pause’ the talks only five days after they had started.

On 12 February 2016, the ISSG announced that its members agreed to establish two task forces under the banner of the UN and co-chaired by the U.S. and Russia to secure the delivery of humanitarian aid and to implement a nationwide ceasefire excluding attacks on IS.³⁰ On 22 February Russian Foreign Minister Sergey Lavrov and US Secretary of State John Kerry signed the Syria Cessation of Hostilities Accord to implement a nationwide cessation of hostilities, endorsed by UNSCR 2268.³¹ Subsequently, De Mistura announced that the Intra-Syrian Talks would reconvene on 14 March. He attempted to use the ISSG’s two task forces to facilitate negotiations on the thorny issue of the political transition.³² While De Mistura claimed that a window of opportunity emerged,³³ at the end of the second round of indirect talks there was no breakthrough on any issue including Assad’s fate in a transitionary process. Yet on 24 March De Mistura released a statement wherein he identified 12 points of commonalities.³⁴

The third round of intra-Syrian Talks, held between 13 April to 27 April 2016 focused on the ‘political transition, on governance, and constitutional principles’.³⁵ Yet in contrast with the mediator’s stated optimism about progress made, again disagreement reigned with regards to what such a transition would exactly entail.³⁶ Furthermore, the talks soon became overshadowed by the worsening humanitarian situation in the country caused by regime forces and their allies expanding military operations. Signalling its frustration over these developments and the UN’s failure to address them, the HNC pulled out.

As the Intra-Syrian Talks had again reached an impasse, De Mistura called on Moscow and Washington to work together to revive the cessation of hostilities and foster a momentum conducive for the peace talks. Subsequently, in early May 2016 the cessation of hostilities was expanded to include Aleppo,³⁷ and on 17 May the ISSG found agreement on a comprehensive ceasefire throughout Syria as well as renewed commitment to unfettered provision of humanitarian aid.³⁸ However, the fighting continued unabated, especially as regime and Russian forces stepped up their campaign in and around Aleppo. This prompted De Mistura for several months to keep deferring a next round of talks. In early September, Lavrov and Kerry negotiated the details of a nationwide ceasefire between opposition and government forces to start on 12 September. Yet the ceasefire failed to hold, and on 3 October the U.S. formally suspended the Vienna talks with Russia, accusing Russia of ‘failing to live up to its own commitments’.³⁹

While the ISSG fell back into disarray due to the heightened tensions between the U.S. and Russia, the latter approached Turkey to help broker a ceasefire in Syria. On 28 December 2016 the two countries announced that a nationwide ceasefire was to start two days later. The UNSC on 31 December adopted Resolution 2336, expressing support to the ceasefire. One month later, Russia, Turkey, and Iran held talks on Syria in Astana, Kazakhstan. Held between 23 and 25 January 2017, the Astana talks were presented as dealing with the details of monitoring the ceasefire in Syria and establishing a trilateral body consisting of Iran, Turkey, and Russia to carry out this task. The UN, although invited, was not in charge of the talks and, by the mediator’s own admission, its participation was limited to an advisory role if so requested by the organizers of Astana.⁴⁰ On 31 January the UNSC issued a press statement supporting the Astana process as well as De Mistura’s plans to reconvene a fourth round of the Intra-Syrian Talks.⁴¹ The latter took place in Geneva from 23 February to 3 March 2017. Yet with respect to the political track, the talks only produced an agreement on the three agenda issues or ‘baskets’ – governance, constitution, elections – already outlined in Resolution 2254 while adding ‘counterterrorism’ as a fourth item, per the Syrian regime’s request.

Despite De Mistura’s assertions that ‘we got some agreement on substance’,⁴² the HNC denied they made any such agreement.⁴³ In contrast, in Astana more progress was made but limited to military issues, such as adding Iran as guarantor to the cease-fire monitoring board and, in early May, announcing four ‘de-escalation zones’. Russia and Turkey failed to agree on the borders of these de-escalation areas and on ways to monitor them. For their part, four more rounds of the Intra-Syrian Talks, until mid-December 2017, worked on the four ‘baskets’ but with no tangible result. De Mistura maintained his upbeat assessment of the talks.⁴⁴ Meanwhile, Russia assisted the regime’s territorial advances and Turkey expanded its military operations against Kurdish groups as it began to occupy much of northern Syria.

In January 2018 parallel talks building on the Astana process resulted in the Sochi Agreement. In it, Russia, Turkey and Iran accepted that the UN oversee the establishment of a Constitutional Committee, a body drafting a new constitution in Syria. The UN was to select a list of 50 members to join the committee together with 50 members nominated by the opposition and 50 by the government. De Mistura embraced the agreement as it promised to resuscitate the UN’s role in political talks. However, the Syrian government not only disagreed with establishing a new constitution, it also rejected the UN’s role in providing a list of 50 Committee members.⁴⁵ As a result, no progress was made on constituting the Committee.

In mid-October 2018, De Mistura announced his decision to step down, citing ‘purely personal reasons’.⁴⁶ He singled out the final month of his mediation – November – to ‘verify’ if there was political will to establish the Constitutional Committee.⁴⁷ Yet the Syrian government held on to its refusal of the UN’s role. In his final address to the UNSC, De Mistura apologized for what he could not achieve. He concluded with a brief allusion to the divisions in the UNSC, albeit then thanking the UNSC for maintaining one consistent and unified ‘line’ of support for his mission.⁴⁸

The mediator’s agency

To assess the mediator’s agency during De Mistura’s time in office, our investigation focuses on the mediator’s main initiatives following from our discussion above: 1) the Action Plan for Syria, 2) the Geneva Consultations, 3) the intra-Syrian Talks, 4) the Astana Process, and 5) the mediator’s resignation. Below we unpack how De Mistura shaped each of these core pillars that constituted the UN’s mediation process during his period as mediator. For establishing the mediator’s agency in each we contend with alternative perspectives stressing path dependency or structural and contextual constraints, or that view the mediator as a unitary rational actor.

The action plan for Syria

The Action Plan was De Mistura’s earliest mediation initiative. Three key factors underline De Mistura’s agency over the Action Plan: 1) the decision to embrace local ceasefires despite their contentious history in Syria, 2) the selection of Aleppo as the primary site to launch the initiative, 3) his overall management of the initiative.

De Mistura’s branding of the initiative as a ‘freeze’ to differentiate it from past ceasefires signals his awareness of the controversial features of local ceasefires in Syria.⁴⁹ De Mistura was not only distinguishing the Aleppo Freeze from previous nationwide ceasefires but also from other local ceasefires like the one in Homs.⁵⁰

De Mistura’s effort to support a local ceasefire initiative stands out from his predecessors’ attempts to organize nationwide ceasefires and prioritize the political process. The variance between De Mistura’s approach and his predecessors displays the agency of the mediator and helps rule out the impact of path dependency on constraining his agency. Proposals for the UN to apply a ‘bottom-up’ approach around local ceasefires also emerged before De Mistura’s entry.⁵¹ Yet, notably, Brahimi did not embrace these calls in the way his successor would come to. Shortly after resigning, Brahimi described a February 2014 local ceasefire in Homs as ‘ … part of the government’s war plan … ’⁵² While Brahimi did incorporate negotiations on local ceasefires in Homs during the Geneva II process, he did so alongside the political process to keep talks from stalling. The Homs Ceasefire, however, was not an initiative launched by Brahimi nor was it considered a template for future initiatives. After resigning, when asked about the precedent set by the Homs Ceasefire, Brahimi pointed at the regime having instrumentalized it to force the rebels to surrender and, accordingly, reaffirmed it had not been ‘part of a peaceful solution, it is part of the war solution’.⁵³

Brahimi’s post-resignation reflections of the Homs Ceasefire contradict a conceivable suggestion that the Aleppo Freeze resulted from a sense of path dependency. They also echo the underlying concern surrounding localized truces: in the absence of a credible political process, they resulted in a military victory for the government.⁵⁴ De Mistura’s decision to champion the idea of local ceasefires despite their controversy ultimately falls on him. At one level, the difference between De Mistura’s embrace of local ceasefires and Brahimi’s critical reflections on the legacy of the Homs Ceasefire, counters conceivable assertions that mediators are socialized in terms of how they are selected – that they all have similar backgrounds and share similar key outlooks and strategic perceptions even before joining. On another, the differences in policy positions between the mediator and his staff, would downplay suggestions that his decisions resulted from bureaucratic factors or pressures from within his own organization. Finally, the controversy over the ceasefires leaves little room to argue that the mediator’s hand was forced or his decision to embrace such an approach was determined by the context. In fact, De Mistura underscored his input into the conception of the initiative as he tried to present the Aleppo Freeze as an innovation of his mission.⁵⁵

Interviews with staff in the Special Envoy’s team suggest that the conceptual design of the Aleppo Freeze was strongly contested by some technical experts, thus weakening framings of the mediator as a unitary rational actor driven by dynamics from within the organization. According to several members of the Envoy’s team, the mediator was well informed and cautioned about the history of local ceasefires in Syria resulting in military victories for the regime.⁵⁶ De Mistura clashed with staff who advised that working with local ceasefires required a comprehensive understanding of the roots behind the controversy surrounding them as they would impede any such initiative.⁵⁷ Accordingly, the mediator was informed that the UN’s office in Syria had been previously approached by vulnerable communities looking to break to sieges, but it was the government that did not want the UN involved in mediating any negotiations on the matter as it would have been more heavily monitored and require a human rights framework.⁵⁸ Expanding on the mediator’s resistance to this information, one UN official recalled a clash between De Mistura and some seasoned staff members who in vain tried to explain that leveraging humanitarian aid to persuade the Syrian government to accept ceasefire agreements, as De Mistura believed, was not going to work due to the government’s tested indifference to human deprivation.⁵⁹

Other staff expressed having similar conversations with the mediator, all underscoring that they believed De Mistura entered his post with ‘ready-made ideas’ concerning the freeze.⁶⁰ This reading was confirmed when in February 2015, five months after the plan for the Aleppo Freeze was announced, De Mistura proposed to dispatch a preparatory team to Aleppo to look into the plan’s modalities, accentuating concerns that he had come up with the idea without involving his own staff and indeed without much preparation altogether.⁶¹

Another element that demarks De Mistura’s input into the Aleppo Freeze, was the selection of Aleppo itself as the primary site to launch the initiative. The mediator repeatedly rehashed a set of generic criteria in his numerous defences for choosing Aleppo, often citing the city’s historical legacy, cultural significance, and its symbolic value.⁶² On a more tactical level, De Mistura argued that an alleged ‘stalemate’ reached between the opposition and the regime at the time was a key strategic reason for choosing Aleppo.⁶³ If De Mistura was hinting at a potential mutually hurting stalemate, the evidence does not match his reading of the situation. De Mistura pointed to the presence of IS forces near Aleppo as an incentive that would ‘trigger’ cooperation between the warring parties.⁶⁴ Each of these tenants appeared to be based on his personal assessments rather than an accurate evaluation of dynamics in Aleppo or the advice of experts within his team. Most notably, Aleppo was not in a ‘stalemate’ as De Mistura portrayed. This highlights how it was the mediator’s own and mistaken perception of the context that determined the selection of Aleppo. As of November 2014, government forces had encircled opposition held areas to cut off supply lines. The selection of Aleppo appears to have caught the Assad regime off guard as it rightly assessed that it enjoyed a military advantage in the city.⁶⁵

De Mistura’s rationale for selecting Aleppo over other locations seemed to revolve around his own tactical arguments. For example, his primary proposition that the warring Syrian parties would unite against IS or that a truce would incentivize them to redirect their attention towards the group were based on his own speculation not shared, informed, let alone enacted by anyone or anything else. Indeed, the mediator did not present any compelling evidence or specific examples of consultations he had where opposition and government forces in Aleppo were willing to form any temporary alliances against IS with the government forces that were preparing to besiege the city. Neither side was even consulted about the selection of the city. Nor were there any signals that the opposition felt more threatened by IS than the government. If anything, senior voices in the opposition were raising concern that a military focus on IS drew attention away from a political solution and the removal of Assad from power, which was their ultimate priority.⁶⁶

The mediator’s agency regarding the Aleppo Freeze was similarly suggested during its attempted implementation. He actively sought to make it the focus of his mediation efforts, eliding the aforementioned concerns of the opposition and critics including among his own staff. It was De Mistura who ensured what he regarded as ‘the only game in town’⁶⁷; not the UNSC, the UNSG, belligerents, or the context itself. The choice to negotiate the Aleppo Freeze was based on the mistaken perception of the mediator that negotiating a local truce in Aleppo would prevent rather than reinforce the capitulation of the city and the suffering of its residents. In the process, De Mistura ignored growing evidence, and those reminding him of such evidence, that the regime’s campaign in Aleppo was key to its strategy to prevail by military means. In fact, while continuing to advocate the Aleppo Freeze amidst the regime’s onslaught onto the city and other opposition strongholds, De Mistura appeared to explicitly abet the regime’s agenda when he declared that Assad had to be ‘part of the solution’.⁶⁸ On that same day, reports emerged of the regime dropping barrel bombs in Douma, killing scores of civilians.⁶⁹ His comment, against the backdrop of repeated diplomatic mishaps from the mediator, alienated the opposition so significantly that they declared that they would refuse meeting with the mediator and his team.⁷⁰ This served to signal their rejection of the Aleppo Freeze. Regardless, the mediator continued to be upbeat in his assessment of the regime’s intentions when in a briefing to the Security Council on 17 February 2015 he maintained that they indicated to him a willingness to halt all bombing activities”.⁷¹ The latter was never announced, and the bombing continued.

Meanwhile, De Mistura had yet to set a date for the Freeze. Neither had he clarified any substantial details of his Freeze plans. In a media stakeout following his briefing to the UNSC, De Mistura suddenly downplayed his expectations for the Freeze and added that before setting a date he would first send a UN preparatory team to Aleppo.⁷² While such never materialized, De Mistura followed through with his proposal, meeting with the Syrian government in Damascus in early March, although the opposition had already rejected his initiative.⁷³ On April 9th, Ban Ki-moon put the final nail in the coffin of the initiative, publicly requesting that the mediator re-launch the political process.⁷⁴ Such an intervention from the UNSG had never occurred with Annan or Brahimi. In short, the mediator’s continuation with the Freeze did not appear to be contextually determined. Rather it was fought for by the mediator who appeared unwilling to give in to pressures to abandon it. That the UNSG at long last recalled De Mistura’s flagship iniative, as the mediator considered the initiative himself,⁷⁵ would push back against any framing of De Mistura as a unitary rational actor given the difference between the mediator’s confidence in his initiative and the lack of it by the Secretary General. Only after the mediator unsuccessfully pursued his initiative for months on end, did the UNSG instruct him to change course. It points up to the mediator’s considerable margin of manoeuvre to which the UN’s hierachy only put a stop when persistent failure and embarrassment pressed it to do so.

The Geneva consultations

De Mistura’s agency in shaping the Geneva Consultations and the establishment of advisory boards for civil society during the later intra-Syrian Talks can be identified in its conception, negotiation, and implementation. Conceptually, the more inclusive approach was part of the mediator’s penchant to ‘listen’ to more actors until a perceived opportunity for a political process could be identified. In light of Ban Ki Moon’s push to end the Action Plan and call for a return to the political process, the preparatory nature of the Geneva Consultations points to the leeway De Mistura had in interpreting requests from the UNSG. Rather than immediately work towards a high-level political conference, like the Geneva I or Geneva II peace talks organized by his predecessors, De Mistura engaged in a loose series of meetings with local stakeholders. By organizing more than 230 rounds of meetings with diverse actors, De Mistura was broadening the range of actors and widening the scope of issues to discuss, such as constitutional reform, rather than focusing on the polarizing yet critical issue of Assad’s future.

Further demonstrating the agency of the mediator in shaping the initiative is the variance between his more inclusive approach to negotiating with civil society and the more exclusive approach in this respect of Annan and Brahimi. More explicitly, De Mistura’s bottom-up approach is distinct in that he was negotiating directly with actors outside the government-opposition binary and doing so in a process designed to fulfil that very objective. Unlike a formative political conference, set in a short-time period, the loose and often ad-hoc style of these meetings allowed for multiple stakeholders to communicate directly with the mediator rather than negotiate with one another.

The mediator continued with this approach to Syrian civil society by establishing advisory boards in the later intra-Syrian talks. This evidences the influence of De Mistura on the conceptual underpinnings of the initiative and in broadening the UN’s engagement strategy with different actors in the mediation process. Thus, at the conceptual stage, the mediator’s agency stands out as more plausibly accounting for his initiatives than perspectives stressing path dependency, structural constraints or instructions supposedly coming from within the UN. There were no directives from the UNSG or the UNSC that the mediator should take on such an inclusive engagement strategy. There also were no existing mechanisms left by his predecessors that tied the mediator or limited his margin of manoeuvre. De Mistura’s openness towards civil society actors in both the Geneva Consultations and later with the establishment of advisory boards, also points to the flexibility UN mediators appear to have in selecting which parties to include in mediation initiatives and in designing the respective processes or dialogue mechanisms to engage or exclude them.

With regards to De Mistura’s management of the advisory boards, the mediator’s agency also stands out. Most notably, key advisory boards like the Women’s Advisory Boards, received particular criticism due to the selection of divisive figures and the absence of due diligence.⁷⁶ These criticisms and allegations never dissipated due to poor transparency over the process by which experts in these advisory boards were selected and confusion over their purpose and the actual content of their workings. The persistence of the criticism undermined the credibility of the Women’s Advisory Board. It also marked a blind spot in the mediator’s overall ability to clarify the objective of the particular initiative, the safeguards to protect its integrity, and its particular impact on the mediation process.

On a conceptual level, the mediator’s preference for a more bottom-up approach to mediation is well demonstrated in the loose structure of the Geneva Consultations that opened the door for civil society to engage the mediator. In turn, the establishment of the advisory boards further point to the mediator’s apparent openness to engaging new actors, excluded by the earlier top-down peace conferences of his predecessors. On a more operational level, problems with the selection of the actors who made up these boards and the failure to address these issues highlight a pattern in the mediator’s approach where he drew public attention to his initiatives but failed to adequately resolve concerns with them.

The intra-Syrian talks

UNSC Resolution 2254 mandated De Mistura to convene formal negotiations between the Syrian parties and facilitate a Syrian-led process that would draft a new constitution and hold free and fair elections. Notably, it did not specify how the mediator should oversee such a process, thus allowing considerable leeway. De Mistura’s tenure as mediator saw nine rounds of such intra-Syrian talks. The mediator played a central role in the strategic design and structure of the Intra-Syrian talks, the choice and style of negotiations, and through his management of the series of talks.

Similar to the Geneva Consultations, in terms of the strategic design and structure of the talks, De Mistura opted for a series of loose talks rather than a large, international conference. The initiative indicates a consistent approach carried throughout his mediation in which the mediator favoured gradual even if nominal negotiations over hosting a central peace conference on more decisive but central political issues as was done in Geneva I and II. By naming them ‘intra-Syrian Talks’, De Mistura avoided characterizing the conference as Geneva III and allowed it to be more fluid, opening the door for a series of conferences.⁷⁷

The mediator also played a deciding role in the nature and style of negotiations, preferring proximity talks over direct talks. In contrast, while Brahimi employed proximity talks to de-escalate tensions during the Geneva II peace conference, he used them as part of a larger strategy to push for direct talks on polarizing issues.⁷⁸ This never materialized under De Mistura, nor did he seem to intend this. Instead, De Mistura appeared to focus on ‘technical’ issues and, along with proximity talks, followed through by organizing meetings between technical experts and the different Syrian parties in an effort to develop clearer positions on technical matters related to the four baskets: the constitution, governance, elections and combating terrorism.

A final, but critical, indicator of the mediator’s agency with regards to the initiative of the Intra-Syrian Talks is through his overall management of the talks. In what appears to be a consistent feature of his mediation in Syria, De Mistura continuously raised expectations after each conference despite little evidence to support such optimism. For instance, after the failure of the first four rounds of the intra-Syrian Talks, De Mistura maintained that an agreement was reached on the agenda. In response, the opposition’s delegation refuted making any such agreement, forcing the mediator to backtrack and claim that he meant the UN agreed to the agenda.⁷⁹ Even the claims made in the very public reversal fell short of De Mistura’s consistent emphasis throughout the process that the agenda was already set and established in UN Resolution 2254. Again, the raising of unsubstantiated expectations coupled with the backtracking shortly after, are functions of the mediator’s agency in the negotiation process and the execution of the talks.

The management of the intra-Syrian Talks can be assessed by holding it against the benchmarks De Mistura himself set. Most notably, De Mistura announced that the eighth round would focus on key agenda items he had already tasked the Syrian delegations to prepare in the first two rounds – detailed proposals on the decisive issues of the political transition and governance. After seven rounds, the intra-Syrian Talks did not deliver any considerable discussion on these critical issues. To that point, the eighth and ninth round of Intra-Syrian Talks under De Mistura, did not successfully resolve these issues. The unsatisfactory progress made during the intra-Syrian Talks did not match the mediator’s publicly stated optimism about their progress. The combination of public optimism despite minimal progress cannot be simply cast off as an effort to motivate the parties to keep hope alive. By presenting deliverables that he did not in fact deliver, the mediator drew public attention to (and scrutiny of) the failures of the entire initiative. The choice to continue with a total of nine rounds despite such tepid progress was one driven ultimately by De Mistura’s decision-making even to the extent that an unresponsive environment did not cause him to change course. These talks, and their depreciating credibility prompted international actors, notably Iran, Russia, and Turkey, to launch their own process of negotiations.

The Astana process

The Astana Process itself was not conceived or organized by De Mistura. As described earlier and while the UN came to support the initiative, it was primarily a Turkish, Russian, and Iranian led process. It was, however, ultimately De Mistura’s decision to accept to participate alongside the Astana Process, to lend it heavy and consistent public support, and to insist it was ‘complementary’ to the Geneva Process. Interviews with De Mistura’s staff suggest that there was considerable debate over Astana and that the decision of how to proceed was ultimately De Mistura’s. Relatedly, the perception that two processes could complement one another rather than compete, clashed with the UN’s institutional guidelines for mediation and the decisions and warnings against more than one track of mediation issued by his two predecessors.⁸⁰ To be fair, the UNSC welcomed the Astana process, and the new Secretary-General Antonio Guterres also appeared to back the UN’s participation alongside Astana as he shared De Mistura’s emphasis that the two complemented one another.⁸¹

The combination of debates within the UN mediation team and the discrepancy between the mediator’s decision-making and his predecessors as well as institutional guidelines from within the UN point to another aspect of the mediator’s agency. On the one hand, there appeared to be no clear consensus or group think among different UN decision-makers and members of the mediator’s team on how to proceed. On the other, while the mediator appeared to operate under the perception that engaging with the Astana process could strengthen his position, it also went against the organization’s institutional guidelines (vague as they are) that prioritized maintaining a monopoly on the peace-making process.⁸² On both levels, there appears to be limited evidence to suggest that the decision to support and endorse the Astana process was a function of socialization. Moreover, there is a discrepancy between Annan and Brahimi’s strong insistence on maintaining a monopoly on the peacemaking process, and De Mistura’s repeated emphasis that the two processes could complement one another. This highlights the limits of potential arguments that would suggest that such a decision would be made by any UN mediator, supposedly because all are socialized prior to or upon entering their posts, and operate with the same ruling logic and strategic outlook. At a contextual level, the mediator was not forced to attend let alone continuously endorse the Astana Process given its clear erosion of the mediator’s credibility.⁸³

The mediator’s resignation

Mediators’ resignation exemplifies the legitimizing power that comes with their entry into a conflict and participation in a peace-making process. Indeed, scholarship on mediation highlights the particular weight a mediator’s resignation or threat to resign holds on parties who value continued mediation.⁸⁴ The decision to step-down or continue mediating marks one of the most distinctive signs of an individual’s agency.

A central aspect to mediators’ resignation is how they frame the rationale for the decision to step down in order to advance an objective, press stakeholders to concede or loosen their negotiating position, highlight challenges undermining the mediation process, or maximize their leverage over the different parties to the conflict. In mid-October 2018, De Mistura announced that he stepped down as mediator, citing ‘purely personal reasons’; professing his commitment to continue with negotiations on the Constitutional Committee during his remaining days.⁸⁵ De Mistura’s resignation demonstrates these two related aspects of a mediator’s agency – the decision of stepping down itself and how this is framed. Concerning the first, in the announcement of his resignation, De Mistura expressed that the decision to step down was his and linked to familial matters. Regarding the significance of his resignation’s timing, De Mistura appeared to use his final month of mediation to present a tight timeline to finalize negotiations on the Constitutional Committee.⁸⁶

Progress on the Constitutional Committee stalled over the role the UN should play in its organization; a direct consequence of the Astana process having sidelined the UN. Specifically, the Sochi agreement, an offshoot of the Astana process, announced that the make-up of the Constitutional Committee would consist of 50 delegates selected by the government, 50 by the UN (referred to as the Middle Third List), and 50 by the opposition.⁸⁷ De Mistura highlighted the Sochi agreement’s referral of the UN’s role in the Constitutional Committee in what was most likely an effort to dispel criticisms directed towards the waning credibility of the UN’s Geneva Process and concerns that it was supplanted by the Astana Process.⁸⁸ It is therefore significant that it was the UN’s explicit role in selecting the Middle Third List that the Syrian government rejected.⁸⁹

While De Mistura pledged to resolve this roadblock, he left without any resolution on the matter. It is likely, however, that by stressing that political will was all that remained to establish the Constitutional Committee, the mediator was using his resignation to set his final month as a deadline to reach such a goal. That the mediator already declared his resignation and explicitly detached that decision from the conflict, likely took away from any pressure he sought to place on the parties.

De Mistura kept with the ‘minimalist’ approach he maintained throughout his mediation efforts, focusing on a modest objective – the Constitutional Committee – all while staying clear from explicitly criticizing protagonists for the failings of the political process in the way his predecessors had. De Mistura’s emphasis that he was stepping down for ‘purely, personal reasons’ rather than due to the obstacles to peace he was facing differentiates himself sharply from Annan’s resignation. The latter was accompanied by a bold condemnation of the UNSC and the particular intransigence of the regime.⁹⁰ It also contrasts Brahimi’s decision to upon his resignation call out the UNSC for not committing sufficiently to the mediation process and the regime for not genuinely committing to the Geneva Process.⁹¹

Conversely, in his final address to the UNSC, De Mistura vaguely apologized for what he could not achieve and took a far less critical approach than his predecessors. While he briefly noted the divisions in the UNSC, he made a point to thank the UNSC for uniting behind him and supporting his mission.⁹² De Mistura also positively framed developments on the Constitutional Committee, downplaying the government’s refusal to agree on the list of members who would join it and the consequent failure of the committee to materialize.⁹³

Yet the Syrian government blocked the formation of the Constitutional Committee particularly because of the UN’s role in selecting 50 members. The government emphasized agreeing to an alternative list with the Sochi conveners and Astana guarantors – Iran, Turkey, and Russia – rather than with the UN mediator. This also directly challenged the UN’s authority to oversee such a process and reinforced concerns that the UN had not only lost its monopoly on the peace-making process but had been sidelined by the Astana and Sochi processes.

The mediator’s resignation avoided acknowledging clear setbacks or growing challenges to the legitimacy of the UN’s mediation process and instead continued to offer unsubstantiated public reassurances. This marks an important pattern in the mediator’s overall approach to negotiating and implementing policies. Ultimately, in both his decision of when to step down and how to do so, De Mistura displayed significant agency.

Conclusion

Individual mediators’ agency matters for how mediation is shaped and pursued, at least a great deal more than is commonly acknowledged. De Mistura’s period as mediator in the Syria conflict was defined by the Aleppo Freeze, the Geneva Consultations, the intra-Syrian Talks, supporting the Astana Process, and the mediator’s resignation. We developed a simple but effective method whereby process-tracing identified these key initiatives after which we scrutinized them as outputs under study at both a conceptual and operational level, established the mediator’s agency, and assessed proposed and conceivable alternative assessments. Each of these mediation initiatives, from their conception to execution, were ultimately a function of De Mistura’s decision-making. Studied together, these initiatives draw attention to patterns in the mediator’s approach, further signalling the significance of his input into the mediation process. De Mistura repeatedly demonstrated a preference for loose, vague objectives that appeared aimed at keeping the semblance of a ‘process alive’, even if it was clear that this process went nowhere and lost credibility. The mediator’s sustained endorsement of an alleged symbiotic relationship between the Geneva and Astana Process, despite the minimal progress of the intra-Syrian Talks that followed the Astana process, exemplifies the eroding credibility that followed the failure of the mediator’s efforts and his consistent unwillingness to acknowledge these setbacks. On a more operational level, the mediator demonstrated a pattern of rapid decision-making with limited preparation or attention to detail, coupled with a focus on high visibility and public promotion of initiatives with high, even if changing benchmarks. The mediator built high expectations promoting near-accomplishments despite evidence suggesting that his initiatives were failing. In other words, it was not only the failure of these initiatives that undermined the mediator’s credibility – which some may still argue were due to contextual variables – but also his continued active promotion of these failed efforts and refusal to acknowledge their failure.

In our assessment of De Mistura’s agency in UN mediation efforts in Syria we ruled out proposed or conceivable alternative explanations. One such alternative explanation suggests that UN mediators’ behaviour is a function of socialized norms shared by their peers. Yet De Mistura’s career trajectory and professional experience differ significantly from his predecessors, making it unconvincing to suggest a potential bias in their selection. These mediators do not come from the same pool. Nor is there evidence that warrants viewing De Mistura as a unitary rational actor that would make the UN and not the individual mediator the prime mover of its mediation initiatives. De Mistura did not act upon UN instructions, or indeed what his staff advised him or prepared for him. Indeed, there were no such instructions coming from above. The only time when such a directive came was when after many months of failing efforts to promote the Aleppo Freeze UN Secretary General told De Mistura to change course. The cases of the Action Plan suggested sharp disagreements with the mediator’s own technical staff, and the decision to endorse the Astana process appears to have also been one in which there was no bottom-up consensus. Although not a comparative study of the three mediators, our case points to clear divergences in behaviour between the three mediators in the Syrian conflict – particularly when they handled decisive issues like maintaining a monopoly on mediation or operationalizing their resignations. It is in this context that explanations pointing at path dependency also failed to resonate.

Nor do arguments that see contextual factors as determinant in mediation efforts explain De Mistura’s decisions as he pursued his initiatives. As the earlier discussion on raising expectations on loose grounds indicated, the mediator appeared to push against the current. His initiatives, therefore, are unlikely to have been contextually-determined – even if the context contributed to their ultimate failure. Thus, Aleppo was not selected to be the flagship of his Freeze initiative because of its strategically given location but rather because of his subjective and questionable interpretation of the city’s importance. Nor were there contextual factors or indeed evidence that made the mediator seek out to rally against IS in the expectation that this would promote political dialogue between the different stakeholders. If anything, De Mistura overruled repeated concerns that his approach mirrored the regime’s anti-terrorism narrative. More generally, recognizing the mediator’s agency does not take away from the contextual variables that constrained his decision-making. Indeed, the paralysis in the UNSC, Washington’s disengagement, and Russia’s intervention, all limited the capacity of the UN-led peace process to be sustained and go forward, favouring initiatives like Astana’s. UN mediators undoubtedly draw much of their credibility and leverage from the UNSC, and De Mistura was no different in this respect. Yet our case study of his mediation efforts in the Syrian conflict shows that such contextual variables and external endorsement ultimately did not determine what the mediator conceived of in terms of his mediation initiatives and how he pursued them. We therefore maintain that UN mediators like De Mistura display significant agency as decision-makers.

Our approach aims to put agency firmly back into the comparative study of mediation. The degree of agency exposed by individual mediators may vary, as indeed alternative perspectives underscoring, respectively, socialized norms, organizational dynamics, path dependency and contextual constraints may at times prove to have greater validity than shown in our case study on the Syrian conflict. A comparison of the conflicts in which mediators operate could be helpful in explaining such variations in agency, as types of conflict may correlate with variations in mediators’ agency. This ‘structured-focused comparison’ of mediators could find inspiration from such approaches within FPA, such as the study by Cuhadar et.al.⁹⁴ Furthermore, a comparative perspective on mediators could benefit from the literature on the psychology of political elites to shed light on the varying degree to which mediators assert their agency and on the different ways in which they exercise it.⁹⁵ In turn, IR theory and political psychology approaches to studying foreign policy that put individual decision-makers and their personalities at centerstage could build on our effort to methodically and empirically establish the significance and degree of agency of one type of such decision-makers in international affairs. For instance, Leadership Trait Analysis has shown itself impatient with the largely inconclusive theoretical debate whether or not leaders matter, to instead focus entirely on ‘how leadership and the characteristics of leaders influence foreign policy’.⁹⁶ Such impatience is understandable but its key premise for studying personalities in foreign policy may not convince more structuralist inclined scholars to the extent our approach would. More specifically to the study of mediation, comparative perspectives centred on the agency of individual mediators could be complementary to and strengthen policy prescriptive approaches to mediation that document good practices, required skills and styles of mediators but are mostly void of analysis of the agency assumed to underly them. After demonstrating mediators’ leeway and their ability to develop and pursue their own initiatives can their individual idiosyncrasies and decisions be argued to matter.

Whether succeeding or failing, international mediation is bound to have an impact on the lives of people who find themselves caught up in the mass violence and destruction of a civil war. When mediation can be shown to be significantly shaped by mediators’ agency, it follows that individual mediators carry considerable personal responsibility. Many Syrians had an acute sense of this, witnessing, for example, inhabitants of Kafranbel (Idlib governorate), famous for their political cartoons and banners, reproaching on many occasions De Mistura personally for his mediation style by proclamations such as: ‘De Mistura ideas to reality are like windmills to Don Quixote’.⁹⁷ In its biting sarcasm, the phrase recognizes the challenges the mediator faced just as it punctuates his own agency in failing to address them. Within the UN the responsibility of its mediators has been tepidly acknowledged, as it drew up lessons learned, formulated good practices in mediation, developed guidelines for the recruitment of mediators, and generally vowed to ‘professionalize’ its peacemaking activities.⁹⁸ Yet when agency is to be firmly brought back into our understanding of mediation, it is high time that mediators own up to their responsibility, are thoroughly assessed and evaluated for what they do, and are being held to account. As former U.S. mediator Chester Crocker notes with regard to the position mediators occupy: ‘If a job is worth doing, it is worth doing well. That means that someone must be placed in charge, held accountable, given the requisite mandate and resources, and steadily supported, or else replaced’.⁹⁹

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Notes

1. Princen (1992), 40.

2. Bercovitch and Gartner (2008), pp. 20–29.

3. Bercovitch (1986); Bercovitch and Houston (1996); Bercovitch and Houston (2000); Zartman (2000) & (2008).

4. See e.g., Bercovitch and Schneider (2000); Bercovitch and Houston (1996), pp. 25, 148.

5. Bercovitch and Houston (1996); and Bercovitch and Schneider (2000).

6. Waltman (2022); Lanz et al. (2017); Moore (2014); Whatling (2011); Boulle and Alexander (2020); and Bercovitch (2009).

7. Svensson and Wallensteen (2010) Wallensteen and Svensson (2014); Lindgren (2016).

8. Bercovitch and Houston (1996), p. 19; Crocker et al. (1999); Beardsley et al. (2006); Svensson (2009).

9. Bâli and Rana (2017); Hellmüller (2022); Hill (2015).

10. Hill (2015); Hume (1994).

11. Bartu (2014); Hinnebusch and William Zartman (2016); Mancini and Vericat (2016).

12. Singer (1961); Singer (1968); Waltz (2001).

13. Byman and Pollack (2001), pp. 107–146.

14. Kaarbo (2017); Jervis (2013); Hudson and Vore (1995); Carlsnaes (1992).

15. Hermann (2005); Hermann and Sakiev (2011); Preston (2010); Görener and Ucal (2011).

16. See for example, Cuhadar, Kaarbo, Kesgin, and Ozkececi-Taner (2017) and Cuhadar, Kaarbo, Kesgin, and Ozkececi-Taner (2017). For more on process-tracing and the study of decision-making, see Jacobs (2015).

17. Hermann (1980) and Hermann (2005);

18. See for example, Cuhadar, Kaarbo, Kesgin, and Ozkececi‐Taner (2017) and Cuhadar, Kaarbo, Kesgin, and Ozkececi-Taner (2017).

19. Hermann (2001). Hermann and Hermann (1989).

20. Byman and Pollack (2001), 109.

21. See e.g., Singer (1961); Singer (1968); Hudson and Vore (1995); Waltz (2001).

22. Jervis (2013).

23. Hermann (2001).

24. De Mistura (2014a).

25. De Mistura (2014c).

26. Ibid.

27. Skype Interview with Syrian Civil Society Representative. July, 7, 2017.

28. United Nations Secretary-General (2015).

29. Office of the Spokesperson for the Secretary-General (2016).

30. International Syria Support Group (2016a).

31. United Nations Security Council (2016).

32. De Mistura (2016b).

33. De Mistura (2016a).

34. UN Special Envoy for Syria (2016).

35. United Nations Secretary-General (2016).

36. United Nations Secretary-General (2016).

37. Al Arabiya (2016).

38. International Syria Support Group (2016b).

39. Gordon and Kramer (2016).

40. De Mistura and Egeland (2017).

41. United Nations Security Council (2017).

42. De Mistura (2017).

43. De Mistura (2017).

44. United Nations (2017).

45. Reuters (2018).

46. UN News (2018).

47. De Mistura (2018b).

48. De Mistura (2018a).

49. De Mistura (2014d).

50. Humeidi (2014).

51. Turkmani et al. (2014).

52. Chatham House (2014).

53. Parasiliti (2014).

54. Kenner (2014).

55. De Mistura (2014d).

56. Skype Interview with Mouin Rabbani. April, 14, 2017; Phone Interview with UN Official. August 2017; Phone Interview with UN Official. September, 7, 2017.

57. Phone Interview with UN Official. September, 7, 2017; Al Jazeera (2015).

58. Phone Interview with UN Official. September, 7, 2017.

59. Ibid.

60. Skype Interview with UN Official. June, 7, 2017; Skype Interview with UN Official. April 2017.

61. De Mistura (2015).

62. Nassar (2019); Black (2014a).

63. UN News (2015); United Nations (2014a).

64. See for example, United Nations (2014a); De Mistura (2016a).

65. DiGiovanni (2015).

66. Black (2014a).

67. Black (2014b).

68. Shields and Perry (2015).

69. Middle East Eye (2015).

70. Agence France-Presse (2015); DiGiovanni (2015).

71. De Mistura (2015); United Nations (2015b); Halliburton (2015).

72. De Mistura (2015).

73. United Nations (2015a).

74. Halliburton (2015).

75. Nassar (2019).

76. Kapur (2017); Gambale (2016); Al Arabiya (2016); Mahmoud (2016).

77. UN News (2016).

78. Nassar (2019).

79. De Mistura (2017).

80. Ki-Moon and Annan (2012); Al Jazeera (2012). For the UN’s institutional guidelines, see Ki-Moon (2009); United Nations (2012); S/ (2015)/730 2015.

81. UN News (2017).

82. Ki-Moon (2009), 6.

83. Hellmüller (2022).

84. Princen (1992), 56.

85. UN News (2018).

86. De Mistura (2018b).

87. Many opposition groups boycotted the Sochi Peace Talks..

88. De Mistura (2018b).

89. United Nations Security Council (2018).

90. De Mistura (2018a); Annan (2012b); Annan (2012a).

91. Interview with Lakhdar Brahimi. Paris, March, 16, 2018.

92. De Mistura (2018a).

93. Ibid.

94. Cuhadar, Kaarbo, Kesgin, and Ozkececi‐Taner (2017).

95. See Davis and McDermott (2021).

96. Kaarbo (2017), 21.

97. Lucas (2015).

98. Ki-Moon (2009); United Nations (2012); S/ (2015)/730 2015.

99. Crocker et al. (1996), 190.