Cooking Peace?
Authoritative mediators’ formulation
in the Aceh conflict 2004-2005 and the Kosovo conflict 2005-2007
Jan von Schmettow
Master's Thesis
Spring 2021
Many mediators exercise power across borders, facilitating talks, formulating agendas and manipulating interests of hostile parties. However, the problem of how mediators justify their use of power and how the terms of this justification legitimate mediators’ strategic conduct has not been systematically theorized and tested yet. A configurational theory can provide varied combinations of mediator authority types and strategy. Two types, legal-rational authority and expert authority, will be conceptualized in relation to formulation strategy. The theory hypothesizes that an authoritative mediator’s acceptable formulation suppresses strategic bargaining and nurtures principled bargaining, propitious for agenda-based mediation success. The theory will be tested by an empirical puzzle. UN mediation on Kosovo (2005-2007) and NGO mediation on Aceh (2004-2005) have both been conducted by a directive approach but negotiations failed in the former case and succeeded in the latter. While the general co-variation supports the hypothesis from authoritative formulation, tracing the causal mechanism reveals that the theory cannot explain agenda-based mediation success in Aceh. Among other questions, a new puzzle suggests the viability of mediators’ varied speech acts as a fruitful research problem.
This thesis is the result of my independent work. All sources and auxiliary materials are referenced in the bibliography. Factual mistakes and errors of judgment are mine.
1. Introduction ... 3
2. Leverage and the problem of acceptable use of power ... 4
3. A concept of authority in mediation ... 6
4. A configurational theory of mediator authority and strategy ... 7
4.1. Mediator strategy: formulation ... 8
4.2. Legal-rational authority ... 8
4.3. Expert authority ... 11
4.4. Causal mechanism: disputants’ deference and principled negotiations ... 13
i. Negotiations under the rule of war... 13
ii. Disputants’ deference to authoritative formulation ... 14
4.5. Negotiated outcome: agenda-based mediation success & failure ... 15
5. Research design and method ... 16
5.1. Operationalizing the independent variable: authority types & formulation ... 17
5.2. Operationalizing the dependent variable: mediation success & failure ... 19
5.3. How to trace the causal mechanism: deference ... 19
5.4. Case selection ... 20
i. Empirical puzzle ... 20
ii. Focused commonalities & differences ... 21
5.5. Data & Coding ... 22
6. Analysis ... 23
6.1. The United Nations mediate in the Kosovo conflict, 2005-2007 ... 23
i. The UN mediator’s legal-rational authority ... 25
iii. Causal mechanism: stalemate, strategic compliance and opposition... 31
iv. Agenda-based mediation failure ... 36
6.2. The Crisis Management Initiative mediates in the Aceh conflict, 2004-2005 ... 37
i. The Crisis Management Initiative’s expert authority ... 39
ii. The CMI mediator’s strategic conduct: substantive and procedural formulation... 45
iii. Causal mechanism: stalemate, breakthrough and deference ... 47
iv. Agenda-based mediation success ... 54
7. Discussion ... 56
7.1. Comparing analytical results... 56
7.2. Alternative explanations for agenda-based mediation success ... 58
7.3. Internal theoretical limitations and puzzles ... 60
8. Conclusion ... 61
1. Introduction
It is generally accepted in the literature that mediators intervene by exercising power across borders (Avant et al 2010, 2; e.g. Young 2015 [1967]; 35 f.; Bercovitch & Fretter 2004; 14 f.; Skjelsbæk & Fermann 1996, 95 f., 80; Khachikian 2000, 23, 30-32). Particularly outsider mediators like many international organizations (IOs) and non-governmental organizations (NGOs) intervene if their activity is understood within a social order structured by spatially segmented units of governance, i.e. states (Reus-Smit 2013). The latter formally claim exclusive authority over conflict resolution in their territory (Susskind & Babbitt 1992, 35; Richmond 1999, 45). This observation raises the question about mediators’ justification to exercise power. Corresponding to sovereignty, international mediation is only loosely institutionalized in today’s international order (Greig & Diehl 2012, 173-176). Mediation is conventionally defined as voluntary process in which an independent or autonomous, outside party helps hostile parties by non-coercive means to reach a commonly accepted settlement (Bercovitch & Lee 2003, 2; Wallensteen & Svensson 2014, 2; Young 2015 [1967], 35). Compared to legally binding adjudication, a mediation process is conditioned on disputants’ acceptance (Kastner 2015, 73, 88, 105 f.). Furthermore, unlike coercive diplomacy, balancing and hegemonic imposition by an involved actor, the definitional mediator needs to be independent: independence refers here to the mediator’s pursuit of a common interest without resorting to threats of or the actual use of force (Art & Cronin 2007; Young 2015 [1967], 44; Rauchhaus 2006, 216).
Thus, there is a tension between disputants’ voluntariness to pursue non-coercive negotiations and many mediators’ intervention. Indeed, Bercovitch and Houston (2000, 180 f.) note that “[h]ow mediators choose to avail themselves of their powers depends on the parties’ acceptance or resistance to the mediator’s use of power and resources to mediate […].” Such observations have not seen systematic theorizing in the literature on mediator leverage yet.
the mediator’s authority or exceed it in the eyes of the disputants. How do such configurations of authority types and strategies affect the bargaining process and outcome? To theorize a possible configuration, a functional causal argument for the mediator strategy of formulation will be developed: here, the mediator classifies and sequences issues, and formulates rules and principles governing talks. Two attending authority types will be theorized, legal-rational authority and expert authority, and how they render varied forms of formulation (in)acceptable in light of their respective social purposes.
These theoretical considerations motivate the hypothesis that an authoritative mediator’s
acceptable formulation is conducive to agenda-based mediation success. Testing this
hypothesis, the theory will be applied to an empirical puzzle: the same mediator strategy, a directive approach, is said to have been applied by the UN mediator in negotiations on the Kosovo conflict in the former Yugoslavia (2005-2007) and by the NGO mediator in negotiations on the Aceh conflict in Indonesia (2004-2005) – with different results. The former mediation failed while the latter succeeded. A disciplined configurative case study will be employed to test whether the variation in outcome can be explained by varied authority types justifying the mediators’ respective intervention and strategic conduct. Thereby, the case study tries to gain empirical insight into the use of power also by so-called “weak” mediators.
2. Leverage and the problem of acceptable use of power
There are a wealth of studies at least vaguely touching upon the link between mediator acceptability and strategy (see for helpful reviews with varied foci, Duursma 2014; Vuković 2014; Landwehr 2019; Hellmüller et al 2020). The following discussion is limited to recent developments in mediation studies on leverage. They pose the principal interlocutors for a conceptualization of mediators’ use of authoritative power.
conceptualization fits to diplomats’ “relational thinking” (Adler-Nissen 2015, 286), it does not easily lent itself to ask which use of power (strategy) is acceptable to hostile parties and which is not (cf. Kleiboer 1996, 371).
The leverage concept tends to foster circular theories and tautological explanations. For instance, Beardsley (2009, 279-282) argues that disputants accept weak mediators because they do not upset the balance of forces. But he conceptualizes mediators as “weak” (leverage), who facilitate ceasefire negotiations (strategy). This arguably self-confirms his theory in that facilitation actually does not upset the balance of forces (cf. ibid, 274). Operationalizing weak mediators, members of the P-5 or G-8 are excluded, NGOs and IOs are included, except for the UN (ibid, 286). Weak mediators thus comprise the EU, Mexico, Brazil, South Africa, Australia, Indonesia and India, a country treated as great power in Beardsley’s empirical example (ibid, 282). It can be doubted whether these mediators are accepted because parties expect them to merely facilitate.
Timothy Sisk employs leverage/strategy in a comparative case study (2009, 54-56). His section on the Sri Lanka conflict for instance is titled “Norwegian mediation: good intentions, little leverage”. Consistent with the concept, Norway’s mediation is analyzed in terms of facilitation (strategy). Since conceptually Norway does not have a strategic choice (little leverage), its failure is a foregone conclusion given high conflict intensity. However, it could be asked whether more leverage would have allowed Norway to be accepted at all and if failure could also be explained by Norway’s inability to remain impartial as member of a monitoring committee (Sisk 2009, 160; Höglund & Svensson 2009).
Siniša Vuković (2015) develops a concept of “legitimate power”. Its effective use by IOs is conditioned on disputants’ consent, delivering relative benefits and that the mediator’s “involvement is morally justifiable, given its particular institutional position, integrity, and ability to promote principles of equity, reciprocity, and responsibility” (ibid, 428). This allows also IOs to pursue manipulation, understood as prescribing behavior to disputants (ibid, 428). Vuković conceptual differentiation between conditions for power and the use of power helps tackling the problem of parties’ acceptance of strategic conduct.
this distinction between power/leverage and strategy is a pre-condition for understanding effects of different forms of power through mediator strategy on the bargaining process and outcome. Even though, Vuković’ concept opens up a path to the theoretical problem raised above. Instead of social power and leverage the following opts for approaching this question through a mediator’s authoritative power.
3. A concept of authority in mediation
A conventional approach to authority is Max Weber’s (1978 [1925]). This concept is chosen because much of the literature referenced here builds on it as well. Furthermore, specifically the UN’s own “peacebuilding sensibilities” can be understood in Weberian terms (Goetze 2017, ch. 5; see below).
authority. But precisely for this reason authoritative mediators will struggle for disputants’ belief in the legitimacy of their power. Devious objectives do not invalidate authority per se. A rub of Weber’s concept is that especially a government, which represents a state’s claim to sovereignty, is unlikely to accept commands by a mediator. Scholars have extended Weber’s concept though. They have theorized authority types which rely on alternative directive speech acts and even assertions. As to the former Onuf explicates (2012 [1989], 87):
Directive speech acts also present the hearer with a speaker’s intention as to some act the speaker would like to have performed. Typical directive verbs are: ask, command, demand, permit, caution.
For instance, IOs usually do not command but “request” states to consider X by a moderate directive (Zürn 2017, 265). On the other hand, assertive speech acts are utterances which state that X counts as Y (Onuf 2012 [1989], 84). Assertions are commonly used by “moral authorities” and expert authorities to influence others’ actions: Amnesty International for instance documents and publishes (asserts) human rights violations (Hopgood 200). Insofar as its assertion is accepted as authoritative (truth), e.g. due to the addressee’s legal obligations, these statements can induce behavioral change. This distinction between directive and assertive speech acts hence enables conceptualizing varied authority types, including expert authority as will be seen.
In sum, the following understands a mediator’s use of power (strategy) in terms of authority: if mediation is a non-coercive, voluntary process but also an intervention and therefore in need of legitimation, than the definitional mediator is strongly pressed to claim authority by a social purpose for his/her strategy which is conveyed by directive or assertive speech acts (cf. Natorski 2018, 280).
4. A configurational theory of mediator authority and strategy
The following summarizes the causal argument fleshed out in the next sections:
Independent variable: (In-)acceptable configuration of mediator-authority-type and formulation
Causal mechanism: Disputants’ deference (or opposition) to authoritative formulation
Dependent variable: Agenda-based mediation success (failure)
4.1. Mediator strategy: formulation
To develop a configurational independent variable, the paper focuses on one mediator strategy: formulation. This kind of strategic conduct can be analytically divided into a function to establish and change the negotiation procedure (or form) and a function to influence the
substance of the negotiated conflict.
Procedural formulation comprises agenda management, that is, the including, excluding, classifying and sequencing of issues (cf. Albin & Young 2012, 41). This can make common interests salient for hostile parties (Young 2015 [1967], 54; Schelling 1980, 30 f.). It also involves introducing procedural rules. Setting a deadline also impacts on the dynamic of concession making.
Substantive formulation refers to a mediator’s compromise proposal integrating hostile parties’ interests. It also includes enunciation of substantive principles, like jus cogens norms such as the crime of ethnic cleansing. By enunciating them, a mediator defines limits to military as well as political positions, since the pattern of warfare and negotiations oftentimes lacks a self-justifying natural boundary (Schelling 1976, 153-157).
Finally, a mediator’s proposal of a case-specific formula is akin to a substantive principle governing specific negotiations: it represents a momentary and abstract ground for further negotiations. It thus also affects procedure. A well-known example is the land-for-peace-formula underpinning Israeli-Egyptian negotiations.
4.2. Legal-rational authority
such authority to justify their actions have to preserve disinterestedness as well as impartiality (ibid, 21): ‘it is the rule that governs, not the person’. This is central to the meaning of ‘rational’ in legal-rational authority. In the case of mediators in armed conflicts, legal-rational authority usually derives from holding an office in a bureaucracy constituted by international law. As regards mediation in armed conflicts, its agents typically are officials of an IO, like the UN (cf. Bercovitch & Fretter 2004, 28). IOs are legally constituted by charters which oftentimes envision their role as mediators.
What is the social purpose justifying such interventions by the UN’s legal-rational authority? To boot, since the UN Charter embodies “constitutive values” of “international society”, the UN’s legal-rational authority in armed intrastate conflicts is structured by “a principle of state sovereignty” (Reus-Smit 1997). This principle is closely related to criteria for the recognition of (sovereign) states: during and after decolonization, this changing criterion has been the collective right to self-determination.
According to Reus-Smit’s convincing argument (2013, 151 f.), delegitimation of Empire as form of government, the emerging international law of human rights and the revival of self-determination were intimately connected. After World War II, the anti-colonialists claimed a collective right to self-determination (ibid 177-187). Crucially, self-determination was argued to be the ticket to realize an institutional framework (the state and international treaties) in which human rights could be protected in the first place (ibid 188). Their politics inter alia led to the influential UN Declaration on the Granting of Independence to Colonial Countries and Peoples (UNGA Resolution 1514) of 1960.
(Fabry 2019, 42). Except for the recognition of Bangladesh in 1971, secessionist claims even by de facto states were not recognized by a critical mass of UN members, if such claims were denied by the parent state (ibid).
Thus, the frame of the UN’s legal-rational social purpose is to guarantee sovereignty and self-determination, conditioned on territorial integrity, for the purpose of protecting fundamental human rights and political participation (cf. Cohen 2012, 202-204). Even though this complex purpose is historically changing and may entail contradictive policies of course, it is a constant in the international community’s recognition practice (cf. Barnett & Finnemore 2004, e.g. chapter 5).
Which formulation is enabled by this social purpose of international legal-rational authority? First, the social purpose of realizing sovereignty and self-determination, heightens their saliency irrespective of what the mediator does. Due to the additional claim to govern in terms of impersonal rules, a UN mediator virtually embodies this principle. Governments and their non-state opponents normally know that the mere presence of a UN mediator reconfigures the their “legal asymmetry”, conferring a political voice on non-state actors (Aggestam 2002, 71; Convergne 2016b, 152). An ideal-typical IO is thus not authorized to initiate an own proposal and judgment involving the application of the standard of self-determination in the intrastate conflict at hand (Young 2015 [1967], 55 f.; Khachikian 2000, 15). Formulation will not be accepted by all conflict parties if it is perceived to prejudge the bargaining outcome of a conflict over territorial integrity and self-determination. By contrast, procedural formulation, including agenda management, will be accepted by disputants if it does not touch upon the sovereignty question. Finally, impartiality is constitutive for legal-rational mediators’ conduct: legality implies impartiality. Disputants’ criticism of bias therefore signifies escalation of negotiations and will potentially derail mediation.
Proposition 1: Substantive formulation by legal-rational authority will be opposed by the
disputants and lead to agenda-based mediation failure.
Proposition 2: Procedural formulation by legal-rational authority will be accepted by the
4.3. Expert authority
Authority flowing from expertise rests on the assumption that “knowledge is unequally distributed” and that those who have it should be complied with (Zürn 2017, 271). Apart from this hunch of being rule by knowledge, it is a matter of controversy how expertise is to be conceptualized. There are a fundamentally different approaches to “expertise”, embedding the concept in varied methodological and ontological contexts (see for reviews Bueger 2014; Allan 2018; e.g. Leander & Wæver 2018). Following Hopgood here (2009, 238 f.), expertise is not merely marked by a claim to speak in the name of ‘impersonal’ and ‘impartial’ international law but to speak in “the absence of any interest or worldly identity – precisely the ‘view from nowhere’” (Hopgood 2009, 238 f). If this claim to represent things how they are constitutes stabilized authority in mediation, conflict parties’ deference is constituted by their outward or genuine recognition of mediators’ objective knowledge (Zürn 2017, 271). The following draws on David Kennedy’s concept of expertise as “rule by articulation” since it elaborates on this point. Conceptually, the aspiration to objectivity in expertise is also central to at first sight incommensurate approaches like Tetlock’s positivist one (2017 [2006], 10-19).
For Kennedy, experts rhetorically position themselves outside of politics – its ideologies, distributional struggles and particular interests – speaking in the name of a “common-sense picture of the world” like a world of states, a global economic order or an Earth System (Lövbrand et al. 2009). This entity requires policy in terms of “the public good, the general will, the practical necessities of reason or the objective truth of scientific knowledge” (Kennedy 2016, 3). Experts thus do not see themselves as deciding on gains and losses but as advising, interpreting, informing (ibid., 116-119). They typically govern by assertive speech acts rather than by directive ones (ibid. 135). Drawing on extensive research of humanitarians, development specialists and his participant observation as legal expert, Kennedy argues (2016, 145):
The basic unit of expert assertion is a link between a proposal about what to do, a reason, and an outcome. There are two steps here: an argument about what a foundational fact, theory, interest, or ideological com-mitment suggests ought to be done and an argument about the outcome that can be expected to be achieved. When I trained military commanders in Africa for the US Navy, we made lots of arguments in this form. In kinetic operations, you must ensure your use of force is proportional to its objective: in this situation, that means this bomb rather than that one. Why? Because international humanitarian laws—and your coalition partners if you want to play in the big leagues—require it. What to expect? If you do this, your campaign will be ethical, legitimate, and more effective and your military will be able to interoperate with ours. Com-pliance is a force multiplier.
bello. Rather, the assertion utilizes law as an instrument or tool to solve a given, extra-legal
problem: how to effectively cooperate with military allies. Consequently, expertise does not consist, for instance, in the weighing of precedents and of applicable legal rules from a bounded corpus to determine the law – involving legal means and ends. It most likely involves legal, moral or other propositional content, but this is not the assertion’s point. Rather, expertise is characterized by a statement about the facts on the ground and about how to solve the problem – understood in terms of legal, scientific, military strategic or moral materials. In sum, one can plausibly conceptualize expertise in mediation as asserting instruction-type rules in the form “If X then Y” (Onuf 1994, 11). Within expertise, instructions rest on an accepted assertion of how things are in themselves, an attendant problem, and a causal argument about probable (or necessary) consequences of adhering or not adhering to them.
Of course, not every expert is an authority in the sense of being the superior in a hierarchical relationship. By reference to which social purpose does an expert claim such a position? To put it simply, expert authority legitimates itself by providing guidance to fix a problem – both in the senses of ‘fixing its meaning’ in relation to a common-sense picture and of “managing” this problem (Kennedy 2016, 95-99). For instance, after the financial crash in 2008, economists could maintain their tarnished expert status by promoting a very specific kind of problem analysis (Kessler 2016): it was a “market failure” defined in terms of information asymmetry and lack of accountability. The crisis was e.g. not widely understood as failure to provide for public participation or as failure to promote distributional justice (ibid. 2016). Defining the problem in terms of the technical knowledge and tools which experts themselves can bring to bear on it constitutes their position of authority.
Self-conscious examples of attempts to develop expertise for mediation are provided by the UN Secretariat’s Mediation Support Unit (MSU), by the NGO-led Mediation Support Network (MSN), by the European Union’s Mediation Support Team and by the African Union (see Convergne 2016a; 2016b; Kastner 2015, 124-126; Kastner 2021, 20; Lehti 2018, 108 ff.). The MSU tries to “professionalize” mediation. It develops standards to train, certify and evaluate mediators, thereby increasing the UN’s authority to govern mediation as a “specialized activity” (Convergne 2016b, 185 f.). A good example is the MSU’s influential UN Guidance for Effective
From these traits of expert authority one can infer kinds of formulation probably acceptable to disputants. Expert authority allows for procedural formulation by assertive speech acts (not by directive speech acts). Differing from the procedural focus of legal-rational authority, expert authority also allows the mediator to make substantive assertions about what will work and what will not work. Most generally and forcefully, an instruction from expertise can take the form of ‘If you want to effectively negotiate, to reach an agreement, a compromise, security, welfare, peace etc. then this is what you have to do, if not, then not.’ Thereby, the expert’s assertion turns a non-derogatory and binding legal principle or moral precept into a hypothetical, which is conditional on actors’ choice. In this vein, enunciated principles, such as self-determination, prohibition of ethnic cleansing or human rights become (in)effective “tools” to be used (or not) ‘in the hands’ of expert mediators and disputants (e.g. Hellmüller et al. 2015, 13). Such changing status of rules supports a frequent argument made about and by NGO mediators. Accepting their assertion does not carry the institutional consequences as accepting the same proposal coming from a position of legal-rational authority does (e.g. Convergne 2016b; Shea 2016, 184; Lehti 2018, 126 f.).
That said, like the UN but in a different way, an expert mediator treads a fine line having to appear impartial or even “neutral” as a consequence of his/her view from nowhere. This version of the view from nowhere is virtually ubiquitous in expert mediators’ authority claims (Kastner 2021, 23-25). Their “neutrality” and “impartiality” hence overlaps with but has a different basis than the impersonal, bureaucratic aspiration of legal-rational authority.
Proposition 3: Procedural and substantive formulation by expert authority will be accepted by
disputants, conditioned on assertive speech acts as well as impartiality, and lead to agenda-based mediation success.
4.4. Causal mechanism: disputants’ deference and principled
negotiations
i.
Negotiations under the rule of war
modelled by a mixed-motive game (Young 2015 [1967], 25 ff.; cf. Sebenius 1992, 326-332). It refers to a bargaining situation where some of the disputants’ motives cancel each other out while other motives can only be realized by cooperation.
However, the context of an armed intrastate conflict raises considerable obstacles to reach an agreement (cf. Sebenius 1992, 336): war is waged as a zero-sum game (Fearon 1995, 398). The gains of one side are conceived of as the losses of the other side. For his reason, negotiations following or accompanying armed violence are susceptible to strategic bargaining (Young 2015 [1967], 33). Strategic bargaining stabilizes mutually attributed intentions according to the rule of war: victory/defeat (Richmond 1998, 709; O’Driscoll 2020). Orienting their decisions by victory/defeat, disputants discount the future gains from their relationship (Jervis 1985, 62 f., 67). Put differently, they think their opponent wants to win the war by other means, potentially using a ceasefire or peace agreement to strive for domination (Richmond 1998). Given that decision-making is mutually contingent, attributed expectation and expectation of such instrumental behavior reproduces disputants’ fixation on their future relative bargaining strength (to credibly commit to their position) (Young 2015 [1967], 33, 169; Schelling 1960, 30 f.). In sum, strategic bargaining transfers the military contest of wills into talks, thereby directing negotiations away from common interests (Ury & Fisher 1992, 8).
ii.
Disputants’ deference to authoritative formulation
How can mediators alleviate strategic bargaining and cultivate what is called “principled negotiations” (Ury 2007; Fisher & Ury 1991, 10)? How can hostile parties’ orientation along the rule of war be transformed?
In procedural formulation, a mediator shapes, organizes and puts into a temporary order the issues to be discussed. Such classification renders incompatible positions salient as consisting of a class of problems including but going beyond the militaries’ battlefield maps (Schelling 1960, 58-67; cf. Barnett & Finnemore 2004, 31 f.). Acknowledging and understanding each other’s positions is required for identifying a common interest lying behind these (Fisher & Ury 1991, 16-18). Ideally, a novel classification of the intractability enables dovetailing differing interests or enlarging the pie (Fisher & Ury 1991, 25 f., 39-41).
principles provide shared means to re-interpret and evaluate each other’s position, behavior and events within a non-military frame. In addition, substantive principles such as self-determination/sovereignty, human rights or democracy can be used to justify concessions
vis-à-vis disputants’ principal and to save face (Young [2015] 1967; Goddard 2012). In this way,
substantive principles politically enable other-regarding proposals and disputants’ concessions during negotiations (Fisher & Ury 1991, 45).
Furthermore, a mediator’s authoritative procedural rule/principle can undermine the rule of war
as such. A good example is the classic balance-of-power principle. It required to take others’
(power-political) interests into account since any claim and offer had to be shown to not upset the balance of power and lead to hegemony (Osiander 1994). Given authority of such a procedural principle, disputants can disclose their interest and reformulate their position regarding the other’s interest, since they expect the other to do so as well. Thus, irrespective of persisting psychological distrust, an authoritative procedural principle can enable opponents to negotiate a shared problem.
4.5. Negotiated outcome: agenda-based mediation success & failure
Defining “mediation success/failure” has been notoriously difficult (Kleiboer 1996, 361 f.; Svensson 2014). Kleiboer has noted that definitions of mediation success are underpinned by scholars’ theoretically embedded normative outlook, whether explicated or not. To alleviate the tendency to smuggle a normative theory into coding mediation “success/failure”, this study applies Svensson’s and Duursma’s concept of mediation success (2019).
concluded ceasefire or peace agreement (ibid. 307).” These mediation goals enable more fine-grained case comparisons.
5. Research design and method
The theoretically derived hypothesis states that acceptable formulation by an authoritative
mediator is conducive to agenda-based mediation success. The following research design
prepares a “disciplined configurative” case study (George & Bennett 2005, 74): it is meant to guide inquiry of an empirical puzzle by developing a new theory for a relatively unexplored phenomenon. The empirical puzzle tackled here will be discussed in selecting cases. I wager that a theory for the relatively unexplored phenomenon of ‘formulation by an authoritative mediator’ can provide an answer to this puzzle. It will be asked whether and how legal-rational authority and expert authority have variedly interacted with the same strategy of formulation, possibly eliciting varied effects via distinct mediation processes.
The authority concept has the benefit that ‘non-authoritative formulation’ does not simply refer to the ‘absence of authority’ but also means ‘transgression of authority’ potentially eliciting disputants’ opposition to the mediator’s power. Thus, the empirical analysis also seeks to shed some light on how mediation failure is possibly induced by mediators’ transgression of authority and their ‘intrusive’ strategy. This argument serves as counterfactual to a potentially positive effect of authoritative formulation on mediation outcome.
5.1. Operationalizing the independent variable: authority types &
formulation
The first (necessary) indicator to account for the presence of legal-rational authority is whether the mediator is delegated by an international organization, either as a state representative, or as an international civil servant. If so, the historical specificity of the IO’s charter as practiced by past mediators and member states before the mediation attempt needs to be investigated (second
necessary indicator). It would be misleading to account for a UN mediator’s authority in 2005
by e.g. using a definition of sovereignty prevailing during the UN’s colonial trusteeship system. The third (sufficient) indicator are relevant IO resolutions and a mandate. It is sufficient since IO mediators may operate without mandate, such as the UNSG’s good offices do sometimes.
1. Necessary: Is the mediator delegated by an IO?
2. Necessary: What is the meaning of sovereignty/self-determination (social purpose) as practiced by the IO and its member states before the mediation?
3. Sufficient: Does a resolution/mandate specify the mediator’s legal-rational authority? Operationalizing expert authority is trickier. It is not “institutionalized”, e.g. codified in a corpus or charter (Sending 2015). However, it is oftentimes made explicit. Thus, the first
(necessary) indicator for expert authority is a mediator’s own justification for his/her
mediation: it is a claim to be able to solve a problem (social purpose) by means of one’s
expertise in solving such problems. It is based on a view from nowhere (e.g. objective, neutral,
independent) wrapped up in a common-sense picture (e.g. ‘scientific’), which underpins the mediator’s problem-solving knowledge or know-how.
How can we know whether disputants recognize specifically expert authority? The second
(sufficient) indicator for expert authority will be disputants’ explicit acceptance of a mediator’s
expertise. Disputants’ acceptance is signified by their reference to the mediator’s past problem-solving, i.e. to the mediator’s reputation, experience and proof of know-how in previous mediations or negotiations in conflict resolution.
constitute self-conscious sites of the expert network, seeking to create and promulgate mediation expertise as well as to visibly professionalize mediation (Lehti 2018, 106, 109 f.).1
This indicator has the limitation of being merely a causal, “input-based” indicator for expert authority relations (Gerring 2011, 183). Furthermore, it could also be criticized for being biased towards Western civil society: employees of these NGO could be expected to have the financial means, educational background and do not face travel-restrictions to regularly participate at the aforementioned sites of the network (Goetze 2017). In addition, formal entry requirements may prevent NGOs from the Global South network here (Lehti 2018, 111). However, this input-based indicator (extension) should not be conflated with a theoretical property (intension) (Sartori 1970, 1041): being only a sufficient empirical sign, it does not state that there cannot be expert authorities in mediating armed conflict who are not part of this network.
1. Sufficient: Do the disputants accept a mediator from the network of mediation experts? 2. Necessary: Does the mediator claim expert authority in mediation, by referring to her*his experience, lessons learned, best practices and/or know-how in solving the problem (social purpose)? Does the mediator reason by a view from nowhere within a common-sense picture?
3. Necessary: Do the disputants justify their acceptance by reference to the mediator’s expertise, experience, reputation and/or know-how in solving the problem (social purpose)?
Mediator strategies are sometimes coded as mutually exclusive (e.g. Beardsley et al. 2006).
More intrusive strategies override the ‘presence’ of less intrusive ones in this coding scheme. Thereby, potential causal effects of different strategies are isolated – if varied strategies have been applied in a single case. But this theoretical assumption about ‘intrusiveness’ is one of the questions of this paper. Therefore, both substantive and procedural formulation will be coded. Tracing the causal mechanism will unveil their relation to parties’ mode of interaction, strategic bargaining or principled negotiations.
1. Does the mediator try to introduce and/or apply a procedural rule, including sequencing and classification of issues?
1 States and IOs can equally build and mobilize expert authority. However, given that the studied cases concern armed intrastate conflicts
2. Does the mediator try to introduce and/or apply a substantive principle, formula and/or a compromise proposal?
3. Are the above rules introduced by assertive and/or directive speech acts vis-à-vis the disputants?
5.2. Operationalizing the dependent variable: mediation success & failure
As discussed above, mediation success/failure is analyzed here as the mediation outcome relative to the goal of a negotiation’s agenda. The first indicator consists of items on a negotiation’s agenda before the beginning of talks (to be potentially renegotiated). Following Svensson and Duursma (2019, 307), possible goals include “establishing an agenda”, “agreeing on a ceasefire”, “concluding a peace agreement”, and “implementing an agreed upon ceasefire or peace agreement”. Here, the most “ambitious” goal on the agenda is coded as the relevant benchmark for assessing mediation success/failure. On the other hand, the actually mediated outcome of talks is compared to this most ambitious goal of the agenda (second indicator). If it is met the mediated outcome will be coded as mediation success, if not this signifies mediation failure.
These two related indicators for the dependent variable have at least the following limitation: they are related. For all its advantages in increasing validity and researchers’ normative detachment, this concept of ‘mediation success/failure’ violates the time order of a neo-positivist causal argument. An agenda, comprising the first indicator for mediation success, is usually agreed upon before or when the mediator’s authority and strategy (X) are present. This indicator might affect the independent variable and the causal mechanism (deference and opposition to formulation). Indeed, recent studies theorize causal effects of the mediator’s mandate – which usually includes an IO mediator’s agenda and goal (e.g. Nathan 2017). While hanging together, the two indicators can be differentiated of course: the “objective” yardstick of ‘what actually happened’ (in comparison to the previously stated goal) distances the dependent variable from the independent variable.
5.3. How to trace the causal mechanism: deference
principled bargaining and partial escalation (Kriesberg & Dayton 2011, 152 f.). These four
indicators are “output-based” (Gerring 2011, 118 f.). They can be justified because they operationalize a causal mechanism in comparison to a conceptual “container” of data (Sartori 1970).
Deference to the mediator’s formulation is indicated by principled bargaining in accordance
with the mediators’ agenda management, classification, sequencing as well as substantive and procedural principles.
Deference and critique to the mediator can be observed if disputants criticize mediator bias, but
still engage in principled bargaining. If they pursue principled bargaining, then disputants’ critique of a mediator’s bias is likely just honest. Together these two indicators are necessary and sufficient for deference and critique.
Strategic compliance refers to negotiations shaped by parties’ devious objectives and strategic
bargaining as described above. To ‘window-dress’ strategic bargaining, disputants use the mediator’s principles and rules to rephrase their position without altering it: they continue the war by other means. Strategic bargaining is necessary and sufficient for the process being coded as disputants’ strategic compliance. Critique of bias is a sufficient but not a necessary indicator.
Opposition to the mediator’s authority, not merely to his/her specific use of power, will be
operationalized by strategic bargaining, disputants’ critique of the mediator bias as well as by attempts to escalate negotiations. An escalating move consists, for instance, in personal attacks of the other party or the mediator or in a military threat (Kriesberg & Dayton 2011, 152 f.). However, these indicators capture not merely parties’ outward pretense to commit to negotiations, but express direct opposition to the mediator’s authoritative power.
5.4. Case selection
i.
Empirical puzzle
the other hand, Ker-Lindsay (2009), Fridl (2009) and Bergmann (2018) attribute the United Nations’ failure to mediate an agreement between Serbs and Kosovo-Albanians inter alia to such a directive approach. Why did a similar mediation strategy succeed here but fail there? The following empirical analysis will reveal whether an authority theory of mediation strategy can provide an answer to this puzzle.
ii.
Focused commonalities & differences
The two cases fulfill the scope conditions of mediation as defined above. UN mediation in Kosovo was based on a UNSC mandate, prohibiting “any solution that was unilateral in nature, or resulted from the use of force” (in Ker-Lindsay 2009, 27, cf. below). NGO mediation in Aceh came about by an informal process initiated by the Indonesian government: since the rebel group of the Free Aceh Movement (GAM) also declined a previous bilateral agreement with the government shortly before, in October 2004, it was not forced to enter into negotiations (Morfit 2007, 117 f.; 121 f.).
In addition to satisfying the scope conditions for cases of mediation, the UN’s mediation and the CMI’s mediation share theory-relevant or focused commonalities justifying a cross-case comparison (George & Bennett 2005, 70). First, both cases represent armed intra-state conflicts pitting a rebel organization against the government by an internationally recognized state: relatedly, both conflicts center on the incompatibility of clashing claims to self-determination/sovereignty on a contested territory. As argued in the introduction, this asymmetry is probably one of the “typical” problems for acceptable mediation in intrastate wars (Gerring 2004, 346). Secondly, both mediations were embedded within Serbia’s and Indonesia’s transformations, including democratic reforms and, more broadly, a reordering of public administration (Miller 2009, ch. 8; Glenny 2012 [2000], 677-682; Đinđić & Bajić 2018). However, the two cases also feature theory-relevant differences which could impact both on the negotiation process and outcome. The tsunami in Aceh on December 26th 2004 constitutes the
initiated by the Indonesian government and not by the UN.
A further difference concerns the international context (Kleiboer 1996): great- and interested powers have (unsuccessfully) attempted to regulate the Kosovo conflict before and during mediation. This contrasts with the Aceh conflict, which has only seen close international attention after the tsunami of December 26th 2004. Finally, the Indonesian government does not
represent an ethno-nationalist project itself, but the Serbian government does (Ramet 2002, 35-38).
These differences also imply that this is not a “most-similar” research design but a “disciplined configurative case study” (George & Bennett 2005, 75): its goal here is to develop a better explanation for successful mediation in a secessionist conflict: a successful explanation would justify additional tests to ask for “contingent generalizations” for this conflict type (George & Bennett 2005, 110, 114 f.). This can only be decided by testing and discussion of alternative explanations. Alternative theories addressing the differences identified above will be the most relevant.
5.5. Data & Coding
Data on the mediator’s strategic conduct and on talks in general is difficult to obtain (cf. Höglund & Öberg 2011). The best sources available here will be first-hand accounts and recollections by participant observers: this is a problem for a theory based on the form and content of communication. To report something is not the same as saying/doing it, especially if the saying is the doing (as during peace talks). To remain skeptical, it will be asked what a given unit of analyzed data/text could mean instead of confirming the operationalized theoretical category. The unit of analysis will be a paragraph.
In addition, like all first-hand accounts, primary sources are susceptible to backward/forward-looking biases (Clark 2013, 9-13): apart from memory gaps, public actors have an incentive to misrepresent the past in light of their prospective agency in the future. To alleviate this problem, first-hand accounts will be triangulated if possible, i.e. compared to other observers’ independent descriptions of the same action/event (Dulić 2011, 38 f.).
first-hand accounts of talks. Finally, the disputants’ representatives have sometimes created personal recollections. Other valuable first-hand reports could be found in US diplomatic cables published on the WikiLeaks platform.
In the case of UNOSEK’s mediation on Kosovo, UNSC’s resolutions as well as the UNSC’s mandate are data, not merely primary sources on data. In the case of CMI’s mediation on Aceh (Indonesia), the NGO’s own publications and especially the Annual Report 2005 represent primary sources on data for its authority claim. They are retrospective and hence not themselves data.
6. Analysis
6.1. The United Nations mediate in the Kosovo conflict, 2005-2007
The following introduction is not a historical argument but flags the parties’ incompatibilities for the reader. The major incompatibility between the Kosovo Albanian interim administration and the Serbian government in 2005/2006 concerned sovereignty over Kosovo. Thus, the first Kosovo Albanian grievance was the abolishment of Kosovo’s constitutional status of an “autonomous province”. This status of a quasi-republic without right to secession as of 1974 was terminated by the Federal Presidency of Yugoslavia under its new President Slobodan Miloševic (Clark 2013, 281; Sundhaussen 2012, 187-189, 258, 260; Ramet 2002, 35). Kosovo Albanians subsequent declaration of independence by a referendum was not recognized by the international community, except for Albania (Salla 1995, 429; Clark 2013, 287).
and their commitment to “non-violent resistance” (Judah 2008, 69 f., 72-74). In this process, Kosovo Albanians were systemically subjected to arbitrary violence by security forces, ranging from harassment to shootings (HRW 1999, xi). According to Human Rights Watch, the security forces created a daily “state of terror” (ibid).
The third incompatibility arose from mutual atrocities and Serbian ethnic cleansing of Kosovo Albanians during the Kosovo War. From 1996 onwards, the Kosovo Liberation Army (KLA) mounted armed hit-and-run attacks against Serbian forces (Kubo 2010, 1144-1146; Judah 2008, 79). During its counterinsurgency, Serbian police and special forces killed one of the founders of the Kosovo Liberation Army (KLA) and his clan in the village of Prekaz, in the Drenica region in February/March 1998 (Judah 2008, 81). Due to this “rallying cry”, the KLA became a popular armed rebellion (Kubo 2010, 1146). Serbian forces responded by a heavy-armed counterinsurgency: between 200.000 and 300.000 Kosovo Albanian civilians were systematically displaced by Serbian security forces until November 1998 (Kubo 2010, 1148; Malcolm 2006, 147; Gow 2012, 307 f.).
Amidst further violence, diplomatic efforts by the until then largely absent international community pressured the hostile parties into negotiations in Rambouillet (France) in February 1999 (cf. Gow 2012, 308-311; Ejdus 2020, 86). These talks entailed Serbia’s second grievance, this time specifically vis-à-vis the US and NATO. The Contact Group, except Russia, revolved around a coercive diplomacy (Sundhaussen 2012, 383-385; Malcolm 2006, 149 ff.; Pouliot 2010, 196-202). Its ultimatum directed inter alia the retreat of Serbian forces, stationing of NATO troops and alleviated autonomy of the province (similar to its pre-1989 status). While the Kosovo Albanians would have provisionally agreed (until a hoped for referendum on status), the Serbian government rejected the plan: in the view of the Serbian chief negotiator, accepting it would not merely violate Serbia’s territorial integrity but also put the Serbian nation in mortal danger (Eijdus 2020, 86 f.; Malcolm 2006, 150; Glenny 2000 [2012], 657).
The war was ended by negotiations between the Serbian government, Russia, Finland, and the US: their agreement was codified by UNSC Resolution 1244 (discussed below). Under Chapter VII of the Charter, the latter stipulated the deployment of a NATO peacekeeping force (KFOR), with the authority to “enforce” the cease fire by force if necessary, and a transitional UN administrative mission for Kosovo (UNMIK). The fifth year of the ensuing state-building project saw intensifying ethno-nationalist violence among Kosovo Serbs and Kosovo Albanians in March 2004 (HRW 2005): nineteen people were killed during riots. These events informed a review by the UN’s Special Envoy Kai Eide, who recommended a more “dynamic” diplomacy compared to the “standards before status” approach taken by UNMIK (Ker-Lindsay 2009, 20-24). Eide’s report of November 2004 to UNSG Kofi Annan ultimately motivated the UNSC’s mandate for Martti Ahtisaari’s mediation.
i.
The UN mediator’s legal-rational authority
1. Is the mediator delegated by an IO?Special Envoy Martti Ahtisaari, his deputy Albert Rohan, and their team (UNOSEK) were appointed by the UNSG and confirmed by the UNSC in November 2005. In its mandate, the UNSC reproduced the Contact Group’s earlier “guiding principles for the future status process for Kosovo” (UNSC 2005; cf. in Ker-Lindsay 2009, 142-144) – which will be analyzed below.
2. What is the meaning of sovereignty/self-determination (social purpose) as practiced by the IO and its member states before the mediation?
The international community’s practice of recognizing claims to independence of Yugoslavian republics was ad hoc and partially inconsistent. While the UNSG Javier Perez de Cuellar and his SRSG Cyrus Vance emphasized territorial integrity and insisted on a consensual solution, the EC’s member states recognized Croatia, Slovenia and Bosnia-Herzegovina in January 1992 (Fabry 2010, 196 f.). These republics were then recognized by most UN members as well. By contrast, secessionist claimants on these republics’ own territory were rebuffed, including e.g. claims by the ‘Republic of Serbian Krajina’, the ‘Bosnian Serb Republic’ and by the ‘Republic of Kosova’, irrespective of these entities de facto territorial control (Fabry 2019, 43; Caplan 2005, 138).
The EC justified this recognition practice by analogy to the post-colonial principle of uti
possidetis in UNSC Resolution 1514 (1960): in this case it referred to self-determination within
and only in existing republican boundaries (Fabry 2019, 44; Hannum 1993, 38 f., 52 f.; Rich 1993, 43, 48 f.). The criterion of “consent by the parent state” was analogized with Yugoslavia’s “federal dissolution”. Yugoslavia’s dissolution referred to the lack of the republics’ representation in the Yugoslavian Presidency, Serbian obstruction of peace talks and armed ethnic conflicts. Finally, the EC added criteria of adherence to the rule of law, democracy and human rights as conditions for statehood. These reconfigured criteria for recognition were formulated during the early 1990s and e.g. recognition of Croatia was not in accordance with them (Rich 1993, 56).
accords” including promotion of the rule of law, human rights and minority rights, as well as democracy (UNSC 1999, 4). The UNSC thus operationalized the EC’s criteria for sovereign
statehood through UNMIK.
Affirming Serbia’s territorial integrity while mandating UNMIK to build Kosovo’s state was perceived as contradictive policy by the affected parties. According to Crampton (2012, 162),
[a]ny attempt to enforce [UNMIK’s] authority in Serb-majority areas, or to hand competences to the overwhelming Albanian-dominated Provisional Institutions of Self-Government in Pristina, was seen by Serbs as a prelude to moves toward Kosovo independence. Conversely, any compromise with Kosovo Serbs (or more usually directly with Belgrade) was seen by Kosovo Albanians as an erosion of their future sovereignty or, worse still, the prelude to Kosovo’s reintegration with Serbia.
In consequence of UNMIK’s state building then, the very foundation of a prospective UN mediator’s legal-rational authority was unsettled, that is, the meaning of “sovereignty” and “self-determination” (cf. also Knoll 2005; cf. Seymour 2017, 831).
3. Does a mandate specify the mediator’s legal-rational authority?
The legal ambiguities were not clarified and resolved in the UNSC’s mandate to Ahtisaari and his team. As noted above, the Security Council simply reproduced the great powers’ guiding principles (UNSC 2005, 2 f.), demanding inter alia that:
1. The settlement of the Kosovo issue should be fully compatible with international standards of human
rights, democracy and international law and contribute to regional security. […]
2. The settlement should ensure multi-ethnicity that is sustainable in Kosovo. It should provide effective constitutional guarantees and appropriate mechanisms to ensure the implementation of human rights for all citizens in Kosovo and of the rights of members of all Kosovo communities […]. [In addition] [t]he settlement should provide mechanisms to ensure the participation of all Kosovo communities in govern-ment […].
6. The settlement of Kosovo’s status should strengthen regional security and stability. Thus, it will ensure
that Kosovo does not return to the pre-March 1999 situation. Any solution that is unilateral or results from the use of force would be unacceptable. There will be no changes in the current territory of
Ko-sovo, i.e. no partition of Kosovo and no union of Kosovo with any country or part of any country. The territorial integrity and internal stability of regional neighbours will be fully respected. […] [emphasis JS]
Thus, the 1st principle and earlier references anchor the mandate in UNSC Resolution 1244
inter alia. The 2nd principle flags institutional protection of non-Albanian ethnic groups in
Kosovo, mainly Kosovo-Serbian residents. These protections include guaranteeing ‘human rights’ and ‘mechanisms’ for ‘participation’ in Kosovo’s future government.
The 6th principle contains the crux of the status matter. It declares the mediator’s goal to be a
principle’s second sentence demands to prevent a return to the situation before the war began in March 1999. This implies that the UNSC has endowed its mediator (UNOSEK) with the legal-rational authority to rule out any bargaining position envisioning Kosovo as dependent province of the Serbian republic. However, this goal may be understood in at least two different ways: first, it could refer to Kosovo’s autonomy, which had already been enshrined as a goal in UNSC Resolution 1244. Secondly, it could be read as exclusively envisioning Kosovo’s independence – on condition of Serbia’s agreement implied by the subsequent exclusion of a ‘unilateral solution’. After all, the pre-March 1999 history had already seen the ‘solution’ of Kosovo’s autonomy which had been illegally abolished by Serbia and the JNA in 1989 (Sundhaussen 2012, 496 f.). Thus, depending on whether one understands ‘sovereignty’ as Serbia’s territorial integrity or ‘sovereignty’ having been quasi forfeited by the abolishment of Kosovo’s autonomy, discrimination and atrocities during the Kosovo War, interpretations of eligible status solutions will differ.
The 6th principle is more clear in the following. Mobilizing territorial integrity and uti possidetis
juris again, it rules out ‘partition’ of Serb and Albanian territories governed by Serbia and
Kosovo respectively. It also rules out Kosovo’s ‘union’ with Albania. Finally, it meant that a solution could not be found in a territorial swap, for instance in an exchange of Serb-inhabited Mitrovica in Kosovo for the Serbian Presevo Valley inhabited by Albanians (Weller 2008a, 25 f.).
Summary of indicators for legal-rational authority relations
ii.
The UN mediator’s strategic conduct: substantive and procedural
formulation
As argued above, the very presence of a UN mediator ‘enunciates’ the principle of sovereignty/self-determination to negotiations, compared to other mediators such as great powers, NGO, or the Vatican. It is thus unsurprising that UNOSEK was bound to introduce this substantive principle among disputants – also against the mediators’ express will. While the Kosovo Albanians sought independence by self-determination and international recognition as sovereign state, Serbia treated this as violation of its territorial integrity, comprising Kosovo’s territory. This conflicted sovereignty-frame proved to be almost irrepressible during negotiations as the UN mediator Rohan recalls (quoted in Bieber 2009, 102 f.; Perritt 2010, 143).
The UN mediators also applied the principle of sovereignty/self-determination to the disputants’ incompatibility. Ahtisaari’s substantive formulation by assertion was his opening move during pre-negotiations (Ahtisaari 2008, 184; cf. 2009, 44):
I made my first trip to the region at the end of November 2005 and told the leadership in Belgrade that I interpreted the mentioned [6th] principle that Kosovo will not return to the pre-1999 situation to mean
that Kosovo will not return back to Serbia. My hosts, particularly the Prime Minister, did not share this interpretation.
This move responded to what the mediators diagnosed as “Gordian knot” of diametrically opposed and mutually exclusive positions (Rohan 2008, 122). They attempted to solve the parties’ fundamental conflict over the meaning of this principle in this specific case.
UNOSEK also emphasized the “unrecognized politics of peaceful means of Kosovars, which only gradually turned into armed struggle” (ibid). These statements mobilize a reading of the Contact Group’s 6th principle, i.e. they give meaning to the term ‘pre-March 1999 situation’.
Secondly, years of “international administration” by the UN after 1999 had created “wholly new realities” and respective “expectations” by the Kosovo-Albanian populace (Rohan 2008,
ibid). Thirdly, Ahtisaari argued on legal grounds that Kosovo constitutes an exception to the
practice of the self-determination principle – “a sui generis case” – and hence should not be seen as precedent for international customary law (Ahtisaari 2008b, 186). The three arguments undergirded the mediators’ quasi-application of self-determination-as-independence.
Over the course of talks, the UN mediators continued trying to convince Serbia to drop its claim to sovereignty (cf. Weller 2008a, 659). When the first rounds of talks had ended in stalemate, Ahtisaari and Rohan sent a “request” for “realistic and constructive positions” by Serbian negotiators (US liaison officer 2006a; cf. Knoll 2009, 128, fn. 25). This constitutes a moderate directive speech act. In any case, the theory would expect this substantive formulation from legal-rational authority to be opposed by the disputant.
The UN mediators also practiced substantive formulation vis-à-vis Kosovo-Albanians as measured by their responses. UNOSEK repeatedly visited Kosovo to “urge” and “demand” the authorities in Priština to further implement UNMIK’s state-building standards, specifically those concerning protection of the predominantly Serbian minority, property restitution and criminal justice in cases of suspected inter-ethnic violence (US liaison officer 2006b). In this way, Ahtisaari connected UNMIK’s earlier “standards-before-status”-approach to the Kosovo-Albanian bargaining position vis-à-vis Serbs – making statehood conditional on specific domestic reforms. Ahtisaari thereby “called for […] further concessions by Priština on decentralization before status is discussed” (US liaison officer 2006b). This substantive formulation was partially accepted by the Kosovo-Albanian interim administration and negotiators as will be seen. If Ahtisaari could indeed ‘demand’ substantive concessions from the Kosovo-Albanians, this constitutes a successful directive speech act by legal-rational authority. It would contradict the theoretical argument about the constrains of legal-rational formulation.
and holy sites, minority rights, and returns and restitution (Knoll 2005, fn. 23). According to Rohan, decentralization and other “practical” issues were put first to suppress the charged status question (Bieber 2009, 102). In addition, the classification made Serbian interests salient – or so it was assumed: it highlighted competences of Serbian-ruled municipalities, Kosovo Serbs’ rights, the fate of Orthodox sites and the question of Kosovo Serbian refugees and returns (Weller 2008a, 671; Ker-Lindsay 2009, 33; Perritt 2010, 143). Direct talks on Kosovo’s status were envisioned for July 24th 2005 – to be held after hoped for compromises on practical issues.
Following a further attempt of decentralization talks in September 2005, the UN mediators publicly stated that a compromise between the disputants was highly unlikely (Perritt 2010, 160). Accordingly, UNSOEK viewed continuation of negotiations as fairly pointless. On 20th
September 2005 then, the Contact Group mandated Ahtisaari to formulate a substantive compromise proposal (Perritt 2010, 165; Ker-Lindsay 2009, 44). Again employing procedural
formulation, UNOSEK at first did not publish or show its substantive recommendations to the
parties: Ahtisaari decided to postpone publication until after Serb parliamentary elections on January 21st 2007 to not cater to nationalist sentiment (Perritt 2010, 160). On February 2nd 2007,
Ahtisaari only communicated UNOSEK’s interpretation of disputants’ compromise on decentralization, retaining his proposal for Kosovo’s final status (ibid. 163). Following final and inconclusive negotiations, UNOSEK presented the mediators’ proposals including recommendations for Kosovo’s final status to the UNSC on March 26th 2007 (ibid. 165).
iii.
Causal mechanism: stalemate, strategic compliance and
opposition
Against this background, the Serbian government’s subsequent rejection of Ahtisaari’s attempt at authoritative formulation is not surprising. The Serbs simply continued business as usual. Their denial of UNOSEK’s formula thus cannot be solely attributed to a mediator’s overstretched authority claim. However, the Serbs based their rejection on an interpretation of sovereignty-as-territorial-integrity opposing Ahtisaari’s substantive assertion: before the talks, President Tadić stated in the UNSC that (quoted in Reliefweb 2006a)
we remain firmly committed to the fundamental principles and norms of international law; in particular those concerning the sovereignty and territorial integrity of internationally recognized states.
Serbia backed this position by reference to the international community’s past recognition practice, including the EC’s recognition and non-recognition of claims to independence in the former Yugoslavia, as well as by an interpretation of UNSC Resolution 1244 (Ker-Lindsay 2009, 26).
Following the first round of talks in February 2006, Prime-Minister Koštunica argued in a speech for the Serbian parliament (quoted in Reliefweb 2006b):
And before the negotiations actually started, there were people trying to fill the void in tangible actions with, say, ‘creative’ interpretations of the Contact Group principles or, more precisely, by finding what did not exist in them. Their favorite argument has been that “to ensure that Kosovo does not return to the pre-March 1999 situation” means to ensure that Kosovo does not return to Serbia. Not only have they blatantly twisted the Contact Group guiding principles, but also the U.N. Security Council Resolution 1244, even before the status talks got under way.
The Serbian government hence directly resisted the mediator’s attempt to mobilize UNOSEK’s legal-rational authority. After all, the UN mediator’s authority was constituted through the very principle fought over by Serbs and Kosovo Albanians. An argument against UNOSEK’s assertion about sovereignty was, by extension, an argument against UNOSEK itself. Furthermore, Koštunica later also personally criticized Ahtisaari for overstepping his mandate: in his view, Ahtisaari declared Serbia “guilty as a nation” (quoted in Ker-Lindsay 2009, 42). Judging by Ahtisaari’s recollection, this was exactly not what he meant. In any case, Serb strategic bargaining, their rejection of UNOSEK’s constitutive principle and personal critique of the mediator’s use of authority represents opposition to the mediator’s legal-rational authority.
For their part, Kosovo’s interim administration would have probably accepted UNOSEK’s formula. However, their acceptance would not be sufficient to distinguish between genuine
deference and strategic compliance to the mediator’s legal-rational authority:
Kosovo’s administration “[f]rom beginning to end [….] insisted on outright independence. […] And whatever the factions of Kosovo politicians, they were all unified in their unwavering demand for independence” (Weller 2008a, 669). This participant observation by a Kosovo-Albanian legal advisor, sympathetic to their cause, rather indicates their strategic bargaining of the basic incompatibility. It implies strategic compliance to the mediator’s authority as will be seen below.
In January 2006, before the beginning of direct talks, the Serbian and the Kosovo Albanian negotiators both adopted and later published negotiating platforms. Firstly, the Kosovo-Albanian negotiators’ platform envisioned Kosovo’s independence by the end of 2006 (Sundhaussen 2012, 496-498): it indicates strategic bargaining just as well as deference to Ahtisaari’s previous formula. Secondly, the Serbian platform restated its claim to “sovereignty” as “territorial integrity” and offered autonomy to the Kosovo-Albanians, consisting in budgetary control (taxes), less-than-sovereign membership in IOs and qualified relations with other states (Ker-Lindsay 2009 33 ff.; Weller 2008a, 668 f.; Sundhaussen 2012, 496-498; Reliefweb 2006c). It followed Foreign Minister Drašković’ earlier formulation of “more than autonomy, less than independence” (quoted in Knoll 2009, 125). This positioning substantiated non-recognition of the UN mediator’s previous authority-backed formulation. Running alongside the rhetorical alignment with the Serbian Orthodox Church’s uncompromising stance, the platform committed the government still further to its position (Ejdus 2020, 104). These actions indicate further opposition to UNOSEK’s authoritative strategy.
As described above, Ahtisaari and UNOSEK retrieved their legal-rational authority by procedural formulation: on January 12th 2006, UNOSEK sent formal invitations to the
disputants, positing Kosovo’s decentralization as topic for the first round of direct talks (Weller 2008a, 671). These ultimately commenced on February 20th 2006. Outwardly, the disputants
heeded the UN mediators’ classification into practical decentralization and other issues rather than discussing Kosovo’s ultimate legal status.
Strategic compliance during talks on Kosovo’s decentralization
