In Search of the Golden Formula: Trends in Peace Mediation Research and Practice

ABSTRACT

This paper examines how the research on, and the practice of, peace mediation has evolved in the past 25 years, with a particular focus on the hypothesised factors that explain mediation ‘success’ and argues for an explicit re-centring of the political in peacemaking. The analysis highlights how research on peacemaking has seen a growth of quantitative studies, while at the same time the practice field of peace mediation has been characterised by a process of professionalisation. We argue that in parallel to these two trends, there has been a shift away from focusing on exogeneous factors, such as those pertaining to the conflict context, to explain ‘success’ or ‘failure’ towards those endogenous to the peace process. A rapidly growing literature on elements of peace process design ranging from inclusivity, mediator characteristics, mediation styles, as well as the substance of negotiated agreements has both informed and been informed by developments in the practitioner community of mediation. These mutually reinforcing trends, while enriching the field, risk portraying mediation as a technical and de-politicised exercise and create inflated expectations of the role and capacity of mediators. We illustrate these trends with a discussion of the case of UN peacemaking in Yemen.

Introduction

The literature on international peacemaking has grown in both theoretical and methodological diversity in the past 25 years. Alongside this, the practice of peace mediation has become a professionalised field. The scholarly literature on peace mediation can be broadly divided into those with endogenous and exogeneous outlooks. The former turns inwardly into the peace process, examining different elements of process design and mediator styles to explain mediation ‘success’ or ‘failure’,¹ while the latter points to external factors pertaining to the conflict context and includes debates around ‘ripeness’, ‘spoilers’ and ‘the political marketplace’ (Stedman 1997, Zartman 2001, de Waal 2012). While one can observe that earlier research on international mediation, such as the concepts of ripeness and spoilers, had more of an external-looking bent, a number of factors have contributed to an increasing number of studies with an endogenous orientation today.

This paper examines developments in the study and practice of peace mediation, notably the increased focus on endogenous factors of process design, a quantitative turn in methods, and the professionalisation of the practice field of international peacemaking. It highlights how these trends have reinforced one another and in turn led to a tendency to view mediation as a technical exercise, along with an overestimation of mediator capacity and influence. Against the backdrop of today’s rising geopolitical competition, we join the call to temper these inflated expectations (Verjee 2016, pp. 3–4) and instead to re-politicise the understanding and practice of mediation as an inexorable part of international and domestic politics (Convergne 2016, McCulloch and McEvoy 2018).

In recent years, there have been several prominent reviews of international peace mediation research (e.g., Duursma 2014, Wallensteen and Svensson 2014, Hellmüller 2023). Many of these reviews examine the strides that have been made by the diversification of methodological approaches in studying peacemaking, particularly the growth of quantitative studies that aim to tackle key questions about mediation in a more systematic matter. We build on these reviews of scholarship and bring them into conversation with trends in the practice field, notably the institutionalisation of peace mediation and the proliferation of mediation actors. By combining the two dimensions of research and practice, the article contributes to debates on the changing nature of peace processes (Turner and Wählisch 2021). Following this introduction, our analysis highlights the growth of quantitative and largely problem-solving research that focuses on factors endogenous to mediation processes, which holds the promise of generalisable lessons or ‘silver bullets’ for mediation. We argue that this, coupled with the oversupply of mediation actors and their logics of professionalisation, has led to a depoliticisation of the mediation field. The argument is supported by an analysis of mediation-related articles published in 10 leading journals over the past 25 years, followed by a brief illustration of these dynamics through the case of United Nations (UN) peacemaking efforts in Yemen.

Changing Perspectives: Outward to Inward Gaze

Existing research on peacemaking sheds light on a number of factors both internal and external to peace processes to explain their outcomes. These include characteristics such as conflict intensity and type, mediator characteristics, or the timing of mediation (Walton 2014, p. 48). Discussions around external factors have been importantly shaped by the ripeness debate (Zartman 2001, Clayton 2013)² and the concept of spoilers (Stedman 1997, Greenhill and Major 2007, Zahar 2008, Nilsson and Kovacs 2011), as well as the role of regional and international actors (Peceny and Stanley 2001, Wood and Checkel 2013). The notion of the political marketplace also offers a helpful lens to analyse mediation processes in certain contexts, by bringing to the fore the significance of the existing conduct of politics in war-affected countries, within which a mediation process is liable to become subsumed (de Waal 2012, 2014). More recently, scholars have explored the relationships between new trends and dynamics in international politics and mediation; Howard and Stark (2018) and Haspeslagh (2021) examine how today’s proscription regimes can undermine international peacemaking efforts, while Hellmüller (2022) points to macro-level changes, such as increasing geopolitical competition and waning influence of the liberal order, shaping mediation outcomes.

The literature with an endogenous outlook, on the other hand, turns inwardly into the peace process to account for its outcomes. There is a significant body of research on different elements of process design, such as inclusivity (see, e.g., Lanz 2011, Hirblinger and Landau 2020), sequencing (Dorussen et al. 2022), the effects of mandates (Pring 2017, Nathan 2018), and more recently digital tools for mediation (Hirblinger 2022). Scholars have also investigated specific characteristics of mediators – impartial/biased (Svensson 2009, Goetschel 2011); state/international organisations/private (Lehrs 2016, 2022, Shea 2016, Bergmann 2018, Lehti and Lepomäki 2018); Global North/South (Graf and Lanz 2013, Leira 2013, Duursma 2020); as norm entrepreneurs (Hellmüller et al. 2017); their strategies and leverage (Beardsley et al. 2006, Brahimi and Ahmed 2008, Beardsley 2013, Reid 2017), as well as competition among them (Lanz and Gasser 2013, Nathan 2017, Verjee 2020, Lanz 2021). Studies with an endogenous outlook similarly include research on specific contents of peace agreements, such as justice provisions (Baker 2000) and power-sharing arrangements (Hoddie and Hartzell 2010), and the excessive focus on signing agreements at the expense of ensuring their adequate implementation (Hampson 1996, Walter 1999, Fortna 2003). Likewise, research that takes issue with the dominant conflict resolution strategy altogether due to its mechanistic, teleological outlook (Clapham 1998) or its western underpinnings (Salem 1993, Menkhaus 1996, Seul 2021) points to certain inherent aspects of today’s conventional peace processes.

While earlier research on international mediation, such as the concepts of ripeness and spoilers, had a more external-looking bent, the past few decades have seen an increasing number of studies with an endogenous orientation. In order to better ascertain this trend, we conducted a systematic analysis of mediation-related articles published in 10 leading peace and conflict studies journals.³ We searched each journal for articles with the term ‘mediation’ or its derivatives in the abstract published in the 25-year-period between 1998 and 2022.⁴ We then read each abstract to check whether the article studies mediation in the sense of international mediation efforts around violent conflicts, as opposed to other types of mediation or an unrelated use of the term (i.e., X ‘mediating’ Y),⁵ resulting in a sample of 190 articles. We coded these according to their primary methodological approach – qualitative/quantitative – and exogeneous or endogenous focus.⁶ We define as endogenous all issues pertaining to the mediation process itself, such as characteristics of the mediator, their style and strategy, the set-up of the process, including its participants, agenda and other dynamics of the mediation process. We define as exogeneous all issues that are outside the realm of the mediation process itself, such as those related to the conflict, the political context, structural factors pertaining to the international order, or dynamics happening in parallel but not integrated into the mediation process, i.e. military intervention or peacekeeping.⁷

The results show a pronounced difference when comparing the growth of studies with an exogeneous and endogenous orientation. Both classifications of articles steadily increased across the five 5-year-periods analysed from 1998 to 2022. As shown in Figure 1, studies with an endogenous orientation grew at a much higher rate, especially in the 5-year-periods of 2013–2017 and 2018–2022.

Figure 1. Trends in Number of Articles Focusing on Exogenous and Endogenous Factors.

This increase of studies looking primarily inwardly into the process to account for its outcome follows the professionalisation and growth of mediation practice in the 2000s, which we discuss below. This rise in studies on endogenous factors links to growing awareness around the questions of process design, and a corresponding desire to improve mediation practice – privileging the question of how to mediate over the question whether to mediate in a particular case. This is in line with a general tension in peacebuilding research between problem-solving and critical approaches (Cox 1981). While the former seeks to uncover factors that affect the efficacy of mediation or peacebuilding initiatives, the latter instead questions prevailing discourses or ways of thinking and does not ‘accept existing policy parameters as a given or as necessarily legitimate’ (Newman 2009, p. 38). The insight that peacebuilding was being driven by a class of detached international actors with limited understanding of the local realities in conflict-affected societies (Autesserre 2014) added urgency to critical research approaches. These aspire to reflexivity and raise questions about existing institutions, policy assumptions and the interests they serve, for example by asking ‘not whether or not to mediate, but why, and for whom?’ (Rosoux 2022, p. 1733).

Nevertheless, in the fields of mediation and peace processes, scholarship in the past decades has come to be dominated by work that falls on the problem-solving end of this spectrum, in particular the kind employing quantitative methods, whereas critical scholarship has been less prominent, and has generally employed qualitative methods (Hellmüller 2023). Much of the literature on mediation outcomes, for example, is framed as debates – e.g. biased versus unbiased and forcing versus fostering debate – and the common goal of these debates tends to revolve around finding ‘a golden formula on how to mediate’ (Duursma 2014, p. 95). This bears much similarity with the questions around efficiency and success that dominate research on peacebuilding more broadly, often underpinned by the assumption that the failure of international peacebuilding ‘is due to some form of technical or organizational dysfunction that can be fixed by some twists and tweaks’ and that with ‘the right screws to turn … peacebuilding will build peace’ (Goetze 2017, p. 4). Perhaps, the very fact that there is much more research on mediation strategies than its onset (Duursma 2014, p. 95) also points to such epistemological aims in research on peace processes.

It should also be added that, while we focus our analysis primarily on articles in academic journals, several books have importantly shaped the field over the past few decades. Epitomising an exogenous focus on contextual factors, Zartman’s (1989) Ripe for Resolution had a lasting impact by introducing his now ubiquitous ripeness theory. Walter’s (2002) influential Committing to Peace: The Successful Settlement of Civil Wars shifted the focus of conflict management away from classic mediation role towards the need for third-party security guarantees for reaching peace agreements. Twenty years after Zartman’s Ripe for Resolution, Sisk’s (2009) International Mediation in Civil Wars: Bargaining with Bullets exemplifies an endogenous research focus on mediator strategies, finding so-called mediation-with-muscle, or more forceful mediation strategies, to be most successful. Beardsley’s (2011) The Mediation Dilemma adds further nuance to the debate on mediator strategies, particularly leverage, finding that while leverage enables mediators to achieve short-term successes in reaching agreements, it is time-inconsistent due to the artificial incentives introduced by manipulation and is as such not conducive to long-term peace. Another example of the focus on factors pertaining to the mediation process itself can be found in Vuković’s (2016) International Multiparty Mediation and Conflict Management: Challenges of Cooperation and Coordination, which addresses the interplay of several mediators in a given peace process.

A number of edited volumes also deserve mentioning, many of which are also used in professional training courses (about which more below). These include volumes by Touval and Zartman (1985), Bercovitch and Gartner (2010), and more recently, Turner and Wählisch’s (2021) Rethinking Peace Mediation, which takes stock of the past decade of changes in the field of peace mediation, focusing on themes such as professionalisation, the normative turn, and the trend towards inclusive processes. Finally, first-hand accounts by former mediators, members of their teams, or of mediation support actors from the UN, regional organisations, or NGOs find their way into training materials and thus inform the identification of ‘best practices’. This is also mirrored in a number of (auto-)biographies that have enjoyed wide circulation, and which naturally centre the perspective of mediators.⁸

Methodological Diversification: The Quantitative Turn

Much of the early research on peace mediation relied on qualitative approaches, notably in-depth case studies to build theories (e.g. Zartman and Touval 1985, Stedman 1997). It has since grown in methodological diversity, and the past 25 years have seen a rapid expansion of quantitative studies, facilitated by the development of large datasets on mediation (e.g. Regan et al. 2009, DeRouen et al. 2011, Duursma and Gamez 2022). This quantitative turn is also evident in our analysis of the 190 journal articles on mediation published in the past 25 years, where we found a stark difference in the percentage increases of qualitative and quantitative studies. Reflecting the overall growing academic literature on mediation, the number of qualitative articles published in the 10 journals grew by 48.5 per cent from the 10-year-period of 2000–2009 to that of 2010–2019. Comparing the same periods, quantitative studies, on the other hand, grew almost double the rate at 90.9 per cent (see Figure 2). Hellmüller (2023) finds evidence of a more pronounced quantitative turn by weighting the analysis to account for journal ranking, finding that in high-ranking journals, qualitative research has become very rare.⁹

Figure 2. Growth Rates of Qualitative and Quantitative Studies.

Despite the promise of quantitative studies to be able to find the ‘golden formula’, its results have in fact often been contradictory, and have ‘been hampered by problems with data sets, time frames and proxies, and by issues surrounding the general suitability of cross-country quantitative comparison as a tool for examining complex processes heavily determined by the specificities of a particular conflict setting’ (Walton 2014, p. 48). The debate on biased versus unbiased mediators illustrates how qualitative and quantitative studies have been able to build on one other, while also exposing the risk of borrowing each other’s concepts in a slippery manner. Andrew Kydd (2003), through a formal, game theoretical model, argues that a mediator is more effective if they are biased as they are able to convince one side to make concessions (because the information provided about the opponent’s capabilities is more credible); a quantitative study by Svensson (2007) then tests and adds nuance to this argument through its finding that in civil wars biased mediators are only more effective if they are biased towards the government. Yet another quantitative analysis by Rauchhaus (2006) finds that unbiased mediators outperform biased mediators as the former are able to share information credibly. Reflecting on this debate, Duursma (2014, pp. 88–89) points to the lack of conceptual clarity around the term ‘biased mediator’ to account for the confusion. Research on mediator leverage, also one of the most-studied aspects in the literature, similarly suffers from different definitions of leverage and has led to ‘a rather disjointed debate’ (Hellmüller 2023, p. 1859). These debates, by aiming to isolate individual variables linked to mediation ‘success’ or ‘failure’, also struggle to account for the substantial differences in mediator identities that have come to dominate the field in practice, which range not only from weak to strong states but include NGOs, private diplomacy outlets, prominent individuals, regional organisations, and the UN (Whitfield 2010, p. 6).

This highlights not only the problem with coherence in concepts, measurement and data but is also anchored within a problem-solving approach that risks brushing over the need to take a critical and reflective view on data. The proliferation of quantitative research in mediation is a reflection of ‘the quantitative turn’ across the field of conflict and peace more widely (Boesten 2017, p. 508, see also Buhaug et al. 2014, p. 141), and scholars have noted that even with statistical studies, ‘raw data used in the models are often not generated from scratch and are subject to a variety of problems’ (Cheng 2018, p. 291, see also Schrodt 2014). Given that politics permeate numbers (Merry 2016, Cronin-Furman and Krystalli 2021), for quantitative studies to be rigorous, it is essential that one embraces a qualitative approach to data, taking into account the way data is constructed, such as knowing who has collected it, how and for what purpose (Dieckhoff et al. 2018, Dawkins 2020). This is all the more important in fields where research results can have direct policy implications, when policymakers may not have the time and skills to interpret advanced methodological designs in all their nuances.

Professionalisation of Peacemaking Efforts

It is not surprising that the quantitative turn in mediation research has taken place hand-in-hand with the increasing professionalisation of peacemaking efforts and the resultant drive for evidence-based approach that demands quantitative knowledge (Merry 2016). With the mounting critique of international peacekeeping and peacebuilding practices by the 2000s for being ‘too ineffective, too risky and too ambitious’, mediation emerged as a more attractive option that seemed less costly in terms of human life and finances on the part of the international community (Nouwen 2022, p. 35, see also Rosoux 2022, p. 1719). Over the last 30 years, mediation has become a standard, ‘almost automatic’ international response to civil wars (Gowan and Stedman 2018, p. 172),¹⁰ accompanied by a proliferation of willing mediators and investment in training to build expertise, both as individuals and an emerging professionalised field. Today’s mediation field is ‘crowded’ and increasingly competitive and includes state actors, the UN, NGOs, and private individuals, with regional organisations such as the African Union, the European Union (EU) and the Organisation for Security and Co-operation in Europe (OSCE) steadily building their capacities and offering their mediation services (Lanz and Gasser 2013, pp. 1–4); against this backdrop, some peace mediation initiatives are even said to have grown into ‘unwieldy bureaucracies’ (Kastner 2021, p. 33).¹¹ In recent years, non-Western states such as Russia, Qatar, Oman, Turkey, China and South Africa, sometimes referred to as non-traditional mediator states, have offered their mediation services in conflict contexts, at times presenting alternatives to the liberal approach to mediation (Hellmüller 2022, pp. 10–11). We are increasingly seeing several of these different actors positioning themselves as mediators in the same context, for example in Yemen, where UN mediation has in the past been accompanied by involvement of other mediation actors, including the Gulf Cooperation Council and the US (Palik and Rustad 2019), and more recently, Oman (Al-Kharusi and Borck 2021, Worrall, 2021). The ‘supply’ in the international peacemaking ‘market’ has grown exponentially over these years (Lundgren and Svensson 2020, p. 2) along with an increase in donor funding. For example, the annual budget of the UN Mediation Support Unit (MSU), which was established in 2006 to provide professional, cross-cutting support to good offices activities, increased from 12.3 million USD in 2011 to 35.9 million USD in 2020.¹² As a result of such efforts, the field today is increasingly recognised as a distinct profession with specialised knowledge, skillsets, and principles, a departure from being primarily ad hoc and individualised initiatives in the past (Walton 2014, Holper and Kirchhoff 2021, Kastner 2021).

Professionalisation of any field, however, carries certain risks. Sociologically seen as a ‘process to pursue, develop and maintain the closure of the occupational group to maintain practitioners’ own occupational self-interests’ (Evetts 2018, p. 44 in Holper and Kirchhoff 2021), professionalisation in its essence is a self-referential endeavour. Such risks are said to be particularly pronounced in a profession’s pioneering phase and when the drive is primarily voluntary in nature, both dynamics relevant to the mediation field, as ‘an emerging professional identity struggling to assert a still-contested expertise and status might tend to a more self-affirmative rather than client-oriented approach to needs assessment’; under such pressure, the profession is at particular risk ‘of forgetting its external raison d’être’ (Holper and Kirchhoff 2021, p. 361). With a widely accepted need for confidentiality that at times serves as a justification for little oversight, such risks are even more pertinent in the professionalised field of mediation. That these issues exist can be gleaned from the tacit agreement that seemed to have existed among the peacemaking community until recently to ignore the ‘elephant in the room’ whether the field’s professionalisation has actually improved the quality of their work (Holper and Kirchhoff 2021, p. 356).

Scholars have also highlighted how the professionalisation and regulation of the mediation field has not adequately engaged with contextual knowledge and experiences of conflict contexts; ethical standards often reflect western cultural values of individual autonomy and objectivity (Kastner 2021, p. 22), while certain mediation initiatives were guided by western ways of seeing and frames of reference (Harriet Logo 2021, p. 208). In this regard, Nouwen (2022) suggests that the EU’s Concept of Mediation that sets out its ambition as a mediation actor harbours Eurocentric tendencies, a critique that may capture the broader mediation field. While unfortunate, this is not surprising given that much of Goetze’s (2019, p. 352) characterisations of peacebuilding as a socio-political field is true for the sub-field of mediation: a relatively homogeneous group whose ultimate basis for recruitment and career development is a higher education degree from an institution in the Global North, preferably an English-speaking one that is ranked among the top-200 of the world. It is also telling how of the 24 members of the Mediation Support Network, one of the most active ‘global’ networks of primarily non-governmental organisations that support mediation, only 8 are located outside Western Europe and North America even following efforts in recent years to diversify their membership.¹³ The eight organisations focus their activities on their own or regional contexts, while the majority rest of the network predominantly work on contexts outside their own, itself reflecting a deep-rooted hierarchical gaze (Johnson et al. 2022, p. 9).

The Making of Mediation Experts: De-politicisation Through Expertise

The establishment of the UN MSU in 2006 and the publication of the UN Guidance for Effective Mediation in 2012 mark significant moments in the professionalisation of the peacemaking field. Such movements institutionalised mediation within the UN, as well as triggering the growth and professionalisation of the wider field. While consequential in many ways, a particularly pertinent implication of these developments for the purposes of this paper is the commitment to depoliticise mediation as part of its attempt to make it more acceptable to states that were wary of its potential threat to sovereignty (Convergne 2016, Kastner 2021). This was done by emphasising the technical aspects of mediation while underplaying its political nature, and much efforts were dedicated in the subsequent years to making and empowering mediation ‘experts’, many of whom were equipped with expertise in specific thematic areas, such as process design, gender or religion. The ‘agnostic approach’ is clear from the ways in which both the UN and the EU present their mediation-related efforts as technical (McCulloch and McEvoy 2018, p. 478), and as such, ‘non-threatening’ (Convergne 2016, p. 190). In fact, the very ‘presumption that there is a universal agreement on the primacy of a technical approach to mediation’ (Nouwen 2022, p. 38) and that a neutral form of ‘professionalism’ is desirable points to the peacemaking industry’s mirroring of colonial powers’ relief efforts of the past (Peace Direct et al. 2021, p. 20). An outcome of these developments has been that recent approaches to peacemaking increasingly betray a bureaucratic mentality absent of the essence of politics (Chandler 2016).

While the emphasis on expertise can enhance the authority and legitimacy of mediation actors, the professional field has not adequately grappled with the fact that no supposedly technical knowledge or skillsets are neutral or value-free (Haas 2011 in Convergne 2016). The training of mediators, as well as negotiators and key actors in conflict contexts, is a major product in today’s peacemaking industry, and undertaking such a course often serves akin to a rite of passage to gain the title of a mediation expert. It is notable how expertise is associated with an accumulation of generic know-how; the numerous training sessions, workshops and other capacity-building mechanisms, as well as the documentation of lessons learned and best practices, rely on and continue to generate generalisable and often labelled international – in essence, depoliticised – knowledge that can be applied beyond a single context, much rather than messier but more grounded contextual knowledge. There is thus an obvious tension between the recognition of a need for context-specific political knowledge, and the typical ‘training that encourages generic peacemaking skills and a more formalised and technical approach to these issues’ (Walton 2014, p. 57).

The privileging of de-politicised knowledge is reflected in the types of training available in the mediation market today that contribute to the making of experts: training courses offered on specific thematics of mediation – such as national dialogues, digital tools for mediation, local mediation, or on designing inclusion formats – clearly outnumber training courses on specific contextual circumstances, be they political, economic or cultural.¹⁴ This technocratic approach to peacemaking goes hand in hand with the ‘preoccupation with tools’ that leads research endeavours to focus on mediators themselves and the elements they deem they can control, at the expense of the conflicts they seek to resolve (De Coning 2011, p. 117 in Convergne 2016). This is also a characteristic of a supply-driven system; when supply exceeds demand for mediation, as is the case today (Lundgren and Svensson 2020, Aeby and Pring 2023), the mediation actors’ need to mediate becomes linked to its own existence, and may override questions of timing or appropriateness of mediation in a certain context. With the ascendancy of mediation as a frequent response to violent conflicts around the world and the competitive supply market, the focus of the question becomes ‘how to’ mediate, not ‘whether to’ when opportunities arise. It is then unsurprising that research has developed in this direction, with a focus on endogenous factors that mediation actors believe they can control and seek guidance on, rather than on the wider contexts – domestic, regional, and global – of a conflict of concern. With a simultaneous drive towards evidence-based policy and the rigours of donor requirements, quantitative research that can seemingly provide clear(er) answers to certain policy questions, in particular those pertaining to process design and issues that external parties and mediators have some control over, is no doubt a useful asset.

Managing Expectations of Mediators’ Agency: A Reality-Check

Much of existing research on mediation takes mediators as their focus of analysis, often at the expense of conflict parties who are invariably ‘the principal protagonists’ of both conflict and peace processes (Nathan et al. 2018, p. 362). In fact, it has been noted that research on peacemaking has failed to pay significant attention to how the conflict parties perceive peace processes and the mediators, while outsiders’ viewpoints of conflict parties are plenty (Meininghaus 2021, p. 328). Our understanding of the significance of an even wider range of actors, such as entrenched actors with their own interests, similarly remains relatively thin (Wallensteen and Svensson 2014, p. 324). That the mediation literature is currently dominated by an outside perspective with scholars, particularly male scholars, based in western universities producing most high-impact academic knowledge, may partly account for this emphasis (Hellmüller 2023, p. 1861). At the same time, mediators tend to be treated as unitary actors with clear decision-making processes and pre-determined interests and aims;¹⁵ this is particularly true for quantitative studies in which mediators are characterised in a certain way and then treated as an independent variable. Not only does this approach gloss over tensions among mediators, especially pertinent in today’s environment where multiple actors are involved in a single process, but it also overlooks the question of whether a mediator has the capacity and resources to indeed mediate, in operational, political and relational dimensions.

In fact, the extensive literature on different strategies of mediators may lead to inflated expectations about mediators’ capacities to induce outcomes through their chosen strategies. This is not to say that quantitative research systematically over-estimates mediator capacity – on the contrary, researchers go to great lengths to control for such factors in their research designs. Even so, there is a risk of this research impacting the practitioner field in a way that strengthens misguided assumptions about mediators’ roles, simply by the sheer amount of attention given to mediator choices regarding process design and strategies. For example, the forcing versus fostering debate implies that should a mediator decide that a forcing approach would be beneficial in the context, they would be able to manipulate the environment and conflict parties’ calculations to prefer a settlement over continued fighting (Duursma 2014, p. 91). Similar presumption of a mediator’s agency is implied in many studies on mediator traits and process design options (Pring 2022, p. 12). Finally, even discussions about the need for formal coordination in cases where multiple different actors are mediating have a ‘tendency to cast these coordination dilemmas as technical issues, when in fact they are related to emerging tensions between new and established powers in the international political system’ (Walton 2014, p. 57), once again downplaying the deeply political nature of mediation.

What risks being overlooked is the very question of the capacity of mediators. Not only are mediators operating in a sensitive environment where certain parties, especially state actors, may at any point accuse them of interference in internal affairs, but the very nature of mediation means that they are invariably entering into a conflict environment with a complex web of pre-existing interests and power relations. The definitional norms of peace mediation – the norms of consent and protecting lives (Hellmüller et al. 2015) – put armed parties at an advantage, not just over non-armed actors, but also over mediators (Pring 2022, p. 6). In fact, reflecting on his experience supporting the Intergovernmental Authority on Development (IGAD)-led mediation process in South Sudan, Verjee (2016, p. 4) stresses the inherent limitations of mediators and the inflated expectations of them: ‘Mediation, contrary to some opinion, is not about getting what the mediators want. Just because you think something is not the best, or not even a good idea, does not mean you can prevent it or insist on a preferred alternative’. Verjee debunks the myth of imposition of the 2015 agreement, explaining how both the power-sharing arrangement and the push for justice and reconciliation originated from the South Sudanese parties, not the IGAD mediators. Factors that constrain a mediator’s capacity in a given context are many, from macro-level geopolitical developments to the mediator’s position and politics within their own mandate-giving organisation, be they states, international organisations or private actors. A deeper understanding of these essentially political dimensions would also help attribute greater agency to negotiators and other actors in a mediated process.

Illustrative Case: UN Peace Efforts in Yemen

As an illustration of the developments identified above, we provide a short case study of the UN’s mediation efforts in Yemen. We do not claim this case to be fully representative of peace mediation as a whole, nor can we do its complexity justice in a few short paragraphs. However, this brief case summary mirrors the trends described above, adding empirical illustration to the problematic, if unintended, de-politicisation that results from an exaggerated emphasis on factors endogenous to the mediation process and elements putatively under the mediator’s control. The case of Yemen is also emblematic of contemporary peacemaking in several respects. Firstly, while states are still the most common mediator type, the UN is the dominant mediating organisation globally (Greig and Diehl 2012, p. 68). The UN has also been at the forefront of the professionalisation and standardisation efforts described above, which have been emulated by regional organisations and states, all of which have developed guidance documents on various elements of process design and invested in training. Finally, as a fragmented and internationalised intrastate conflict, the conflict in Yemen is characteristic of broader global conflict trends of the past decade (Pettersson et al. 2019).

As a mediator, particularly since the outbreak of the current civil war in 2015, the UN has been operating in a politically difficult environment where the parties show little willingness to settle, partly due to the conflict’s internationalised character (Palik and Rustad 2019). While there have been multiple attempts by the UN - some successful - to reach partial agreements, in the context of an overall stalling of talks, successive UN Special Envoys have focused their efforts on elements they can control, notably consultation mechanisms with Track Two actors and inclusivity questions, such as training of civil society and women’s groups. Meanwhile, UN mediators have been criticised for failing to adapt to changing political realities such as conflict fragmentation, instead holding on to an outdated framework that assumes a two-party conflict between the Yemeni government and the Houthi rebels (Asseburg et al. 2018, p. 56). Furthermore, in the UN’s and the broader peace mediation community’s narrative, the earlier phase of UN involvement in Yemen’s National Dialogue Conference of 2013–14 is continuously hailed as a success case (Asseburg et al. 2018), with reference to its intricate process design elements (Root and Salisbury 2014), including for its reflection of core messages and principles of the UN Guidance for Effective Mediation (Zyck 2014, p. 9). This is despite the fact that the conference’s outcome document was never implemented and a civil war broke out in 2015.

The UN’s good offices were first brought in to support the Yemeni transition in the context of the Arab Spring of 2011. After large anti-regime demonstrations, the Gulf Cooperation Council forged an elite deal that averted large-scale violence and paved the way for President Saleh to step down and for the country to start a process of an inclusive national dialogue, ultimately leading to a new Yemeni constitution. The UN played a key role in the preparation and implementation of the Yemeni National Dialogue Conference (NDC). The UN funded the talks and shaped its preparatory phase, where it was key in promoting the participation of women (Asseburg et al. 2018, p. 51), for which a 30 per cent quota was achieved across all delegations in addition to a separate women’s delegation. The NDC also included intricate participation and voting mechanisms for a broad range of groups including civil society, youth, women, different political parties, and southern representatives. At the conference’s conclusion, it was celebrated as ‘a model for a UN-led transition process’ (Asseburg et al. 2018, p. 52). While the NDC is still described as a positive example in terms of process and set-up, the agreed constitution failed to be adopted and a civil war eventually broke out, initially between the Houthis and the Yemeni government, with other factions emerging, including forces in the South fighting for independence. Critical voices contest the idea that this was a successful model case (Al-Muslimi 2014), since key conflict issues remained unaddressed and the UN was unable to react to changing security developments on the ground (Jalal 2022).

Following the outbreak of war, and in the face of successive failures to achieve progress between the parties, the Office of the UN Special Envoy (OSESGY) has diverted attention to Track Two and Three initiatives, conducting extensive ‘Framework Consultations’ to gather insights on Yemenis’ preferences and concerns, has funded training for civil society in particular for women’s groups, largely conducted outside the country, and has engaged in lessons learned exercises. Funding for some of these activities runs through the UN and donor country-managed Peace Support Facility for Yemen, which aims to ‘deliver discreet activities emerging out of existing Track II initiatives with the goal of catalysing the political process’ (Peace Support Facility 2020). However, activities such as these have had limited impact on conflict resolution efforts and instead have created a self-referential cycle where OSESGY consults, but civil society groups and the UN-formed technical women’s advisory group ‘complain of a disconnect between Track II dialogues and UN-led negotiations’ (International Crisis Group 2020, p. 30). This response to the stalling of the political process is facilitated by trends described in this paper: a focus on process design and a standardisation – if not projectisation – of mediation, leading in this case to widespread disillusionment with the UN-led mediation process.

Conclusion

International peace mediation has developed significantly in the past 25 years, both in its practical application in countless violent conflicts around the world and in its growth as a dedicated field of academic study. Scholars have generated important insights into the factors that enable or constrain mediation in armed conflicts. While these developments are to be welcomed, in this paper we have identified several concerning trends in the field. Based on an analysis of literature on mediation in the past 25 years and illustrated by a short case study of UN peacemaking efforts in Yemen, we have highlighted how the quantitative turn in research and a focus on factors endogenous to mediation processes, coupled with the professionalisation and standardisation of peace mediation practice, risk reinforcing an increasingly de-politicised approach to peacemaking.

Ours is not a linear, strictly speaking causal argument about scholarship impacting practice, or vice-versa. Rather, we understand research foci and practice developments to be mutually reinforcing developments. In practice, we have witnessed many cases where geopolitical realities, such as divisions within the UN Security Council, have necessitated a de-politicised approach to mediation and a focus on putatively technical issues in their stead. At the same time, the large and growing body of literature on elements of process design and mediation strategies, often employing quantitative methods, inevitably buttresses mediation professionals’ desire to identify a ‘golden formula’ for age-old mediation dilemmas, which would respond to the ‘expertise’ requirements of a professionalised sector.

What we argue here is not that scholarship examining the influence of variables surrounding the mediator or mediation process claims to have found the single silver bullet. Nor do we claim that these works have discounted contextual factors and the interests of conflict parties. In fact, scholars regularly and often rigorously include controls for these factors in their research designs. Rather, it is the reception of these studies that concerns us: despite the caveats communicated by scholars, they nonetheless cement an image of certainty and generalisability across myriad conflict contexts that is likely to limit mediators’ capacity to respond to changing developments in specific contexts in a politically astute manner. Knowledge production is thus put to the service of a growing professional community seeking generic tools and know-how, while operating in a situation of over-supply and under-demand. This risks both de-politicising mediation and over-estimating the capacity and role of mediators at the expense of other actors.

Finally, it is important to note that, with the above concerns identified, we by no means imply that mediation should not be pursued. It is in the nature of mediation that it will rarely succeed, given that it is employed in extremely difficult situations where conflict parties are unable to come to agreements on their own. However, ours is a call to reflect on the potential risks of mediators being busy-bodied with things outside their core mandate, diverting attention from politically sensitive to seemingly ‘technical’ tasks. As the saying goes, if the only tool you have is a hammer, you tend to see every problem as a nail: if professional expertise in mediation is focused on process design, then workshops, training, and consultation processes may take centre-stage while failing to reflect on their purpose.

An honest look at the status quo and the above-mentioned self-referential cycle is more important today than ever. In contrast to the early post-Cold War period, conflict resolution through comprehensive peace agreements has become rare in recent years. In a context of increased geopolitical rivalry and frequent conflict fragmentation, coupled with an over-supply of mediation, mediators are in practice increasingly reaching only more modest goals such as local ceasefires or prisoner exchange agreements. At this crossroads following a two-decade long process of professionalisation and standardisation, it is critical for the field to undertake a frank review of its current practices for it remain ‘fit for purpose’ in the years to come. Re-centring the political in peace processes, which involves resisting the urge to identify and design a generic ‘golden formula’, may be a first step in this reckoning.

Disclosure Statement

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

Additional information

Funding

This work was supported by the Swiss National Science Foundation under [Grant number 100017_197543].

Notes on contributors

Jacqui Cho

Jacqui Cho is a PhD fellow with the swisspeace Mediation Programme and a PhD candidate in Political Science at the University of Basel. Her research focuses on peace mediation, DDR, and African politics. She holds an MPhil in African Studies from the University of Cambridge and has worked with various UN bodies in East Africa and New York.

Dana M. Landau

Dana M. Landau is a senior researcher with the swisspeace Mediation Programme and lecturer at the University of Basel. Her research focuses on mediation, peacebuilding, minority rights, and nationalism, and has appeared in International Affairs, Security Dialogue, Journal of Intervention and Statebuilding, and Nationalities Papers. She holds a DPhil from the University of Oxford, Department of Politics and International Relations.

Notes

1. The notions of ‘success’ and ‘failure’ in mediation are themselves contested; see e.g., Kleiboer (1996).

2. See also the Special Issue of Ethnopolitics ‘Revisiting the Ripeness Debate’ (2022): https://www.tandfonline.com/toc/reno20/21/2.

3. These are the Journal of Peace Research, Journal of Conflict Resolution, International Negotiation, International Peacekeeping, Civil Wars, International Affairs, International Organization, International Studies Quarterly, Cooperation and Conflict and Security Dialogue. 18 articles published in Civil Wars were included in the sample. Of the 10 journals analysed, Civil Wars was the journal with the fourth most articles on mediation.

4. For the two journals - International Organization and International Negotiation - whose websites did not allow for a ‘search in abstract’ function, we searched for the terms in titles. The sample includes four review articles.

5. We also excluded articles where mediation is not the focus of analysis.

6. For the latter, we asked the code question, Does the article focus predominantly on factors endogenous or exogeneous to the mediation process as explanations for mediation outcomes?

7. The distinction between endogenous and exogeneous orientations is not always clear cut. To ensure inter-coder reliability, one researcher first coded all the articles, while indicating those that were unclear. The other researcher then blindly coded 20 per cent of the total sample across the selected journals, as well as those marked unclear. The two sets were largely in agreement, and for a few difficult cases, the coders came to an agreement after together examining the individual articles. In a few cases, constituting 8.4 per cent of the sample, the article focuses equally on endogenous and exogeneous factors. In 2.1 per cent of the cases, the focus was found to be on neither. We coded these studies as such – ‘both’ or ‘neither’ – in the Supplementary Material to the article, but exclude them in the analysis that compares the relative growths of the two bodies of research.

8. See, for example, Mitchell’s (2000) Making Peace: Northern Ireland Peace Accord; Holbrooke’s (1999) To End a War; Merikallio and Ruokanen’s (2015) The Mediator: A Biography of Martti Ahtisaari.

9. The most highly-ranked journals in the field, Journal of Peace Research and Journal of Conflict Resolution, publish predominantly research using quantitative methods, whereas for example the lower-ranked International Negotiation is more open to qualitative research and constitutes a larger part of our sample.

10. Studying the period between 1946 and 2010, Diehl et al. (2021) find that mediation has by far been the most frequent conflict management approach in international-civil militarised conflicts.

11. It is worth noting that these activities are sometimes termed ‘facilitation’, rather than ‘mediation’. While some mediators adopt formal distinctions between these activities (see e.g., https://www.eda.admin.ch/eda/en/home/foreign-policy/human-rights/peace/switzerland-s-good-offices/facilitation-and-mediation.html), in practice, these semantic choices are sometimes made in response to political sensitivities or preferences of the conflict parties.

12. We may be seeing a reversal in this trend in recent years, however, with an increase of military and security spending at the expense of donor funding for peace processes. The annual budget of UN MSU struggled to achieve its envisioned 40 million USD for 2022, with only 9.3 million being pledged as of December 2022 (https://dppa.un.org/en/funding). Regional organisations report a similar experience, with the OSCE lacking the necessary political backing and resources to expand its conflict management capacities (Lanz 2021 in Greminger et al. 2021).

13. As of December 2022. https://mediationsupportnet.ethz.ch/member-organizations/.

14. An example of the latter is the Yemen-based Sana’a Center’s offer of an in-depth course on Yemen: https://sanaacenter.org/event/the-eighth-yemen-exchange.

15. Notable exceptions to this include Verjee (2020) and Lanz (2021).