The different shades of EU mediation: A comparative analysis of EU's conflict resolution support for Ukraine and Moldova's unresolved secessionist conflicts (2014-present)

(1)

1

The different shades of EU mediation: A comparative analysis of EU’s conflict

resolution support for Ukraine and Moldova’s unresolved secessionist conflicts

(2014-present)

M.Sc. Crisis and Security Management – Master Thesis

Andrei Stanciu – 2674653

20.06.2020

Thesis Supervisor: Dr. A. L. Dimitrova Second Reader: Dr. H. Mazepus Word Count: 23.938

(2)

2

Abstract

Since the inception of the Concept on Strengthening EU Mediation and Dialogue capacities in 2009 the EU sought to establish a policy basis for its broader engagement in the area of conflict mediation. However, currently very little is known about the empirical manifestation of EU’s mediator roles as described in its Concept document and their effectiveness in positively impacting the various conflict-resolution processes the Union is involved in. Consequently, this thesis examines EU’s mediation involvement in Ukraine and Moldova’s conflict resolution processes, as two relevant examples for its non-direct mediation involvement, and provides an assessment of the necessary conditions for EU’s mediator effectiveness identified as: leverage, strategy and coherence. The analysis reveals that EU’s more effective impact in Moldova’s Transnistrian settlement than in the Ukrainian crisis has been contingent on its greater degree of leverage towards the warring parties and its degree of cohesion with its Member States most relevant to the conflict-resolution context. Conversely, its different mediation strategies have been found to make an impact only to the extent to which they were supported by its degree of leverage and coherence.

Key Words: EU different mediator roles; Ukraine and Moldova conflict resolution; conditions for mediator effectiveness; Concept on Strengthening EU Mediation.

(5)

5

Abbreviations List

AA Association Agreement

CFSP Common Foreign and Security Policy

CIS Commonwealth of Independent States

CPT Causal Process Tracing

CSDP Common Security and Defence Policy

DCFTA Deep and Comprehensive Free Trade Area

DPR/LPR Donetsk/Luhansk People’s Republic

EaP The Eastern Partnership

EBRD European Bank for Reconstruction and

Development

EIB European Investment Bank

ENI European Neighbourhood Instrument

ESS European Security Strategy

EU European Union

EUAM European Union Advisory Mission

EUBAM European Union Border Assistance Mission

EULEX European Union Rule of Law Mission

IBM Integrated Border Management

IcSP Instrument contributing to Stability and Peace

IMF International Monetary Fund

OECD-DAC Organisation for Economic Cooperation and

Development – Development Assistance Committee

OSCE Organisation for Security and Cooperation in

Europe

U-LEAD Ukraine – Local Empowerment, Accountability

and Development Programme

UN United Nations

(6)

6

Chapter 1: Introduction

1.1 Problem Outline

Russia’s annexation of Crimea in 2014, and the consequent armed conflict triggered in the eastern parts of Ukraine, have brought renewed attention towards European Union’s (EU) diplomatic capacities and its foreign policy towards its post-Soviet neighbours engaged in protracted secessionist conflicts (Bogzeanu 2016; Nilsson and Silander 2016; O’Loughlin, Kolossov and Toal 2015). Currently, EU’s commitment to neighbors like Ukraine or Moldova is underpinned by policy doctrines such as good governance promotion or security consolidation. Nonetheless, EU’s approach to these countries, and the Eastern Neighborhood in general, has been highlighted by a grander ambition for stabilization and conflict resolution since the inception of the European Security Strategy in 2003. In this regard, EU has continuously taken a proactive stance towards international mediation, being an ongoing mediator in protracted conflicts such as those of Georgia or Syria. However, EU has also taken upon other roles when it comes to conflict mediation, the Union being currently an observer in the 5+2 peace format over Moldova’s Transnistrian settlement (OSCE 2019) and a direct supporter of the Normandy peace talk format for the Ukrainian crisis, through its member states of Germany and France (EEAS 2015:2). Notwithstanding this fact, research into EU’s different roles in peace negotiations, and its influence on conflict mediation processes, has been indicated by both scholars and analysts to remain a scarcely researched topic (Sheriff and Hauck 2012; Bergmann and Niemann 2015; Visoka and Doyle 2016). Therefore, this thesis’ purpose is to provide a contribution to this research gap and explore EU’s different roles in conflict mediation processes as well as their impact upon the overall conflict resolution context.

1.2 Research Question

Based on the problem outlined above this thesis aims to answer the following research question: Under what conditions did the EU positively contribute to Ukraine and Moldova’s conflict

(7)

7 1.3 Academic and Societal Relevance

During the past decades, international mediation has become a rapidly expanding multilevel and multi-stakeholder peacebuilding affair, leading many scholars to wonder what is the role, or potential, of international actors such as EU in conflict mediation processes (Gündüz and Herbolzheimer 2010:9; Sheriff and Hauck 2012). EU has always been to some extent involved in mediation or dialogue around the world, however, with the inception of the Concept on Strengthening EU Mediation and Dialogue capacities in 2009, the European Council sought to establish a policy basis for its broader engagement in the area of conflict prevention. Yet, scholars (Bergmann et al. 2018:4; Haastrup 2015:3) suggest that currently very little is known about EU’s different mediator roles as described in the 2009 Concept and the extent to which they are reflected by the empirical reality, this providing for a promising new area of study. Moreover, some (Sheriff and Hauck 2012:1; Haastrup 2015:8) suggest that further research on mediation has the potential to increase visibility of this sort of work even within the EU, as in their interviews with certain EU officials they have discovered that the Concept itself, as well as EU’s mediation roles, are not widely known.

In this instance, the choice of Ukraine and Moldova’s secessionist conflicts as a comparative case study of EU’s mediation work is academically relevant for two reasons. First of all, scholars have indicated a dearth of systematic research into both EU’s different mediator roles and their effectiveness (Bergmann et al. 2018:4; Chaban, Elgström and Knodt 2019:519). Thus, a comparative case study research on Ukraine and Moldova has the potential of revealing the different nuances of EU’s mediation, as proposed in its Concept document, whilst process tracing their empirical manifestation through the span of six years can provide a comprehensive overview of the necessary conditions for effectiveness. Second of all, considering EU and Russia’s competing foreign policies in Ukraine and Moldova’s conflict resolution processes, this thesis contributes to an emerging pool of research on the effect international linkages and leverage have on a country’s conflict dynamics (Sasse 2016). Such research has the potential of bridging an apparent disconnect between mediation studies grounded in Conflict Analysis research and EU’s foreign and security policy situated in Security Studies, as Haastrup (2015:2-3) suggests that there is ‘no coherent body of work on EU mediation in the context of its role as foreign policy actor’.

(8)

8

Chapter 2: Body of Knowledge

2.1 Literature Review

EU’s involvement in conflict management and peacebuilding, inside and beyond its neighbourhood, has increased its role significantly since the elaboration of the European Security Strategy (ESS) of 2003. Thus, the institutionalisation of the ESS within EU’s foreign policy, set the Union on a course to promote ‘a ring of well governed countries’ around its troubled Eastern neighbourhood, championing policy doctrines such as good governance, socio-political reform and rule of law (European Council 2003:9-12). This venture into the world of conflict management has been further entrenched with the Concept on Mediation document of 2009, EU conceptualizing mediation as a policy tool for ‘conflict prevention, transformation and resolution […] supported through the full range of civilian and military crisis management instruments as well as trade and development tools available to the EU’ (EU Council 2009:3-4). Therefore, analyses of EU’s peacebuilding practices have generated a variety of theoretical frameworks and conceptualisations, aimed at understanding the full spectrum in which EU seeks, or succeeds, in promoting democracy, security and human rights (Barbe and Kienzle 2007; Biscop 2010; Poggoda et al. 2014; Rabynovich 2019). Broadly speaking, however, EU’s role in conflict mediation builds upon three broad literature strands of EU foreign policy studies.

One such strand concerns EU’s role as global conflict manager and the criteria relevant for assessing its security governance effectiveness within the CFSP/CSDP dimension. Research such as that of Norheim-Martinsen (2010), for example, explored EU’s different security governance patterns, showing how EU’s governance of its CSDP is characterized by features like heterarchy or ideational relations among different actors, indicating a trend towards a supranationalisation of its security policy. Other scholars such as Keukeleire and Raube (2013) have researched CSDP’s governance from a security-development nexus perspective, analysing the extent to which EU’s CSDP signals a shift in EU’s securitization of development foreign policies, showing a limited causation between promotion of security and development governance. Other research into EU’s security dimension referred to the ‘capabilities-expectations gap’ concept, by investigating EU’s response to crises and conflict, Rieker and Blockmans (2019) argued here that EU’s capacity to act remains hampered by limited resources and coherence among Member States despite its

(9)

9 capability of identifying its goals clearly. Supplementary theoretical contributions have also been done towards assessing EU’s military missions’ effectiveness, or its underlying conditions, research such as that of Rodt (2011:42) or Whitman and Wolff (2012:11), pinpointing to indicators such as internal-external goal attainment or capabilities to act and fund, as relevant reference points in conducting empirical analyses.

A second strand of research has been dedicated towards EU’s peacebuilding through integration. Central to this this dimension are EU’s use of positive or negative conditionality, or the externalization of internal governance models, as means to influence a conflict transformation. In this strand scholars such as Diez, Albert and Stetter (2008:26) focused on different pathways through which the EU can foster transformation of border conflicts through integration, distinguishing between compulsory, enabling, connective or constructive policies. Following similar premises, Tocci (2008) and Delcour (2013) researched EU’s impact in promoting peace through the use of contractual relations finding positive causations between EU’s contractual relationships and peace promotion in Turkey’s conflict with the PKK, and Georgia’s one with its secessionist regions. Other theoretical attempts aimed at encapsulating the integrationist peace model employed the neofunctionalist lenses, scholars (Visoka and Doyle 2016; Bergmann and Niemann 2018) theorizing on EU’s peacebuilding model as a spill over effect of previous sectors’ integration with EU’s external policy framework

A third strand of research concerns EU’s actorness and its effectiveness in international institutions or negotiations. Drawing from Sjöstedt’s (cited in Bretherton and Vogler 2006:16) definition of actorness as the ‘capacity to behave actively and deliberately in relation to other actors in the international system’, scholars in this field have been interested in EU’s capacity to act as a fully-fledged independent actor in international relations. The debates accompanying this topic concern EU’s material capabilities and the extent to which it has been able of projecting itself as an influential and independent actor in world affairs, rather than a mere platform for Member States’ interest (Niemann and Bretherton 2013:262). Thomas (2012), in this regard, analysed whether a great degree of EU internal cohesiveness, defined as determinate common policies and their pursuit by EU Member States and institutions, can have an impact on EU’s foreign policy effectiveness. His conclusion was that ‘coherence may be necessary for the EU to exert its influence abroad’, yet not enough of a condition to counteract a multi-centric world where other

(10)

10 actors can deploy more resources to achieve their goals (Thomas 2012:472). Others, such as Vogler and Bretherton (2006:22,) argue in this instance that EU is an ‘actor under construction’ whose capability of projecting itself onto the world is shaped by three underlying conditions of opportunity, presence and capability. Covering various policy areas of EU involvement, their analysis concluded that EU’s range of foreign policies, as well as relationships in the international system, earns the EU the role of model, promoter of values and alternative to the US (Vogler and Bretherton 2006:216-217). Jørgensen, Oberthür and Shahin (2011:603), contribute to this research by theorizing four relevant conditions necessary for EU’s effectiveness within decision-making processes of international organisations, referring instead to effectiveness, relevance, efficiency or financial viability.

Literature on EU’s mediation performance builds on research such as that of the previously mentioned scholars, drawing from the different insights they offer into conceptualizing EU’s behaviour in world politics and its negotiation effectiveness (Bergmann and Niemann 2015:960). That is because while the Concept on EU Mediation document plays an important role in establishing EU’s course of intervention in providing third party assistance to conflicting parties, EU’s conceptualisation of mediation is purposely broad (EU Council 2009:2-3):

Mediation is a way of assisting negotiations between conflict parties and transforming conflicts with the support of an acceptable third party. The general goal of mediation is to enable parties in conflict to reach agreements they find satisfactory and are willing to implement. The specific goals depend on the nature of the conflict and the expectations of the parties and the mediator. A primary goal is often to prevent or end violence through cessation of hostilities or cease fire agreements. In order to ensure peace and stability in the long-term, mediation should be cognisant of and, as appropriate, address the root causes of conflict

Moreover, according to the document (EU Council 2009:6) EU can promote, leverage, support or fund mediation on top of being a direct actor, placing mediation within multiple policy fields of its external relations, providing the opportunity to be deployed at any stage of the conflict cycle in order to prevent or transform it (EU Council 2009:4; Haastrup 2015:3; Bergmann et al. 2018:3). So far, studies on EU mediation have been scarce, most available ones remaining exclusively concentrated on direct types of EU mediation (Grono 2010; Forsberg and Seppo 2010; Natorski 2018). As a result, an argument could be made that most studies overlooked EU’s development of capacities for mediation and dialogue, and consequently, the variety of factors underpinning its

(11)

11 performance as a foreign policy actor in conflict resolution. Thus, the following section will provide an overview of the ways in which this thesis seeks to address this apparent gap.

2.2 Theoretical Framework

2.2.1 Conceptual Foundation

The theoretical framework employed for the analysis of EU’s mediation relates to the one proposed by Bergmann and Niemann (2015:960), as it does not only draw heavily from both international mediation and EU foreign policy literature (Bercovitch et al. 1991:16; Bercovitch 2006:294; Tocci 2008:882-883; Rodt 2011:55; Jørgensen, Oberthür and Shahin 2011:603; Thomas 2012:461), but is intended ‘to be applicable to the full universe of cases of EU mediation'. Bergmann and Niemann’s (2015) framework proposes the analysis of EU’s mediator performance along two dimensions. The first one assesses mediator effectiveness along an internal EU perspective and an external conflict one. The internal EU perspective is concerned with EU’s own ‘goal-attainment’ along three different values ranging from high to medium and to a low degree of effectiveness, where high represents EU’s achievement of most goals it set before negotiations and low entails the achievement of minor goals or none at all. The external conflict perspective is concerned with the observable change in conflict behaviour on the sides of contenders, and distinguishes between six different values of conflict settlement, ranging from full settlement to no agreement.

The second dimension of this framework is concerned with the conditions most relevant to determining EU’s mediation effectiveness. Here, Bergmann and Niemann (2015:961) propose the following four conditions:

(1) Mediator leverage – ‘defined as the resources and instruments the EU can bring to the negotiation table to spur an agreement’

(2) Mediation strategy – ‘strategies EU can adopt to foster an agreement: facilitation, formulation and manipulation’

(3) Coherence – ‘understood as the degree of co-ordination and substantive agreement between individual Member States’ policies towards the mediation activities carried out’ (4) Conflict context – ‘conflict parties’ cohesiveness and willingness to make concessions, followed by the timing of mediation’

(12)

12 2.2.2 Analytical considerations and amendments

The analysis Bergman and Niemann (2015:964) conduct using this framework is argued to represent a ‘plausibility probe’, rather than a causal relationship testing, meant to check the plausibility of a theoretical argument without undertaking an extensive empirical research. Therefore, for this thesis’ research purpose, certain alterations of this model will be made in order ensure analytical feasibility as well as uphold a degree of empirical coherence. These considerations relate both to the dimension measuring mediation effectiveness and the one stipulating the conditions fostering effectiveness. Firstly, relating to the measurement of mediator effectiveness, EU’s internal perspective of goal attainment will be removed from this analysis. That is because interviewing EU officials, as Bergmann and Niemann (2015:964) did, in order obtain indicators of goal-attainment effectiveness goes beyond the capacity and time limit of this thesis’ research. Moreover, the range of EU programming documents characterizing its foreign policies towards Ukraine and Moldova, entail a variety of multi-dimensional policy goals which change on a yearly basis, EU characterizing its latest 20 deliverables for 2020 document, for example, as a ‘flexible living tool’ (European Commission 2017a:1), making the operationalisation of goal achievement indicators impractical and difficult to verify. Conversely, the external conflict perspective effectiveness will be complemented with two additional indicators accommodating the complexities of the mediation environment, as well as EU’s mediation overarching objective which is acknowledged as a long-term endeavour aimed at addressing root causes of conflict (European Council 2009:3). This choice is supported by conflict mediation literature which suggests that assessing effectiveness is contingent upon various factors, the value of mediation going beyond merely producing a peace agreement often making the delineation between success and failure unclear (Kirchhoff 2008:210; Greig and Diehl 2012:105-106). Thus, mediation may change the way conflict parties interact, the foundations of future negotiations; they may improve the humanitarian situation or give hope to affected populations and so forth (Lanz et al. 2008:9). While this might seem problematic for the comparative analysis proposed in this thesis, the purpose of this research is not to conclude on which mediation process is best, but rather, as d’Estrée et al. (2001:102) suggest, inquire into ‘what each mediation process did best’ and what were the underlying conditions determining the specific outcome.

(13)

13 These aspects justify my second major alteration to Bergman and Niemann’s (2015:963) model, which is removing conflict contextfrom the four independent variables determining EU’s mediator effectiveness. That is because the aspects constituting the variable conflict context, such as good or bad timing (depending on conflict intensity, nature of conflict issue, domestic politics, etc.) or conflict parties’ internal cohesiveness (parties’ internal power structure and spoiler problems), are too broadly conceptualized to allow for a feasible operationalisation of indicators which would go in line with the temporal focus of this thesis and its space limitations. That is not to say that the conflict context will be completely excluded from the analysis of EU’s mediation, as Russia’s involvement in both Moldova and Ukraine’s conflicts (Wolff 2012; Moga and Alexeev 2013; Robinson 2016) will be discussed for example when assessing aspects such as EU’s leverage, therefore touching upon the aspect of spoiler problem (representing leaders or parties working against certain mediation outcomes that threaten their power or interests) contained within the conflict context dimension. However, due to considerations regarding this thesis’ analytical feasibility the research focus will be confined to the three dimensions of leverage, strategy and coherence.

2.2.3 Analytical framework and theoretical conceptualisation

The amended analytical framework this thesis proposes consists thereof of a dimension concerned with the conditions fostering effectiveness, specified above as leverage, coherence and strategy, and a dimension concerned with measuring mediator effectiveness as an external conflict perspective (see Table 1, p. 15).

Beginning with the independent variables, mediator ‘leverage’ refers to the resources and instruments (economic incentives, trade agreements, visa liberalisation, etc.) the EU can bring to the negotiation table to spur an agreement. The general hypothesis Bergmann and Niemann (2015:962) propose here is ‘the more resources (leverage) the EU possesses, the more likely it will be an effective mediator’. The ‘strategy’ variable refers to the approach adopted by the EU to foster an agreement. Bergmann and Niemann (2015:962) indicate here three different types of mediation EU can adopt: facilitation (where the mediator serves as communication channel among disputants but does not intervene substantively), formulation (where the mediator takes a proactive stance and shapes the negotiation process) and manipulation (where the mediator contributes to

(14)

14 negotiations but also directly influences the bargaining structure through use of coercive measure or positive incentives). The general hypothesis regarding this variable being that ‘the more “manipulative” the EU’s mediation strategy, the higher chances the mediation effort will lead to a settlement of the conflict’ (Bergmann and Niemann 2015:962). Finally, there is the mediator’s ‘coherence’ which refers to the degree of co-ordination and substantive agreement between individual Member States’ policies towards the mediation activities carried out. Bergmann and Niemann’s (2015:963) hypothesis here is that ‘a highly coherent EU approach towards a particular conflict sends a strong signal of EU’s determination to help the warring parties to solve their conflict and, simultaneously, leaves no room for interpretation concerning EU’s preference regarding the substance of a negotiation solution’.

Regarding the dependent variable, the external conflict perspective dimension proposed by Bergmann and Niemann (2015:961), distinguishing between six different values of conflict settlement ranging from full settlement to no agreement, will be employed. Additionally, two more outcome variables relevant for EU’s transformative role will be included from the Guidance on Evaluating Conflict Prevention and Peacebuilding Activities issued by Lanz et al. (2008) under the OECD-DAC Network, which has been indicated by EU funded research consultants (Gündüz and Herbolzheimer 2010:3; Sheriff and Hauck 2012:1) to provide for a good starting point in evaluating EU’s conflict assistance. Among the nine interlinked criteria for assessment proposed (Lanz et al. 2008:12), only two of them will be used in this research due to their relevance for EU’s work as presented in its foreign policy context, as well as to avoid duplication with other criteria already included. In order to encompass the complexity of mediation environments, Lanz et al. (2018) criteria is structured in a set of open-ended evaluation questions allowing for a systematic and flexible assessment of different aspects. The two criteria of interest to this research will be:  Effectiveness and Impact: assess the changes that an intervention contributed to and whether it has

reached its intended objectives, as well as the different effects of a project or policy – Has the relationship between parties changed as a result of the mediation? (e.g. sense of security, level of violence, humanitarian situation, conflict dynamics, confidence-building measures, etc.)

 Coverage: assesses the inclusion and exclusion of actors, issues and regions – Did the mediation process include (or exclude) the most relevant stakeholders, issues and regions? (e.g. inclusion of civil society, leaders of minority groups, neighbouring states and regional powers)

(15)

(16)

16

Chapter 3: Methodology

3.1 Research Design and Level of Analysis

The research purpose of this thesis is to analyse how the above mentioned causal conditions determine EU’s effectiveness, in its different mediator roles, in the cases of Ukraine and Moldova’s conflict resolution processes with their breakaway regions. Hence, the research design employed will be that of a comparative case study. Case study design is understood in this research according to Gerring (2014:342-343) and Rohfling’s (2012:24) definitions which describe it as an ‘intensive study of a single unit for the purpose of understanding a larger class of similar units’ where that unit connotes ‘a bounded empirical phenomenon that is an instance of a population of similar empirical phenomena’. By that logic, the empirical phenomena concerning this thesis are EU’s different types of mediation, a research attempt which requires a comparison across units.

In order to determine the sort of comparative case study this research is interested in, Rohlfing (2012:12-14) suggests that a reflection upon the level of analysis is needed. As mentioned above, the research goal of this thesis is the ‘assessment of a causal explanation’, meaning that the level of analysis this case study will operate on will be at both cross-case and within-case level. That is because the level of theoretical interest resides partly into investigating Bergmann and Niemann’s (2015) hypothesized causal effect between the three variables and the conflict mediation outcome, as well as uncovering the empirical manifestations of these effects upon the causal process, aspect which is left implicit in their framework. For example, the general assumption regarding leverage is that the more resources the EU possesses, the more likely it will be an effective mediator (Bergmann and Niemann 2015:962). However, as Rohlfing (2012:13) notes, in this situation ‘a hypothesis on the causal mechanism does not necessarily imply an expectation about the causal process’. Leverage in this instance does not specify how EU’s resources lead to mediation effectiveness. Therefore, by analysing EU’s mediation effectiveness based on these conditions I expect this research to (a) develop an explanation on their effect upon the causal process and (b) test the hypothesized effect of these conditions upon the mediation outcome. Thus, according to Rohlfing (2012:15), this sort of binary focus renders the research design to that of an integrative comparative case study.

(17)

17 3.2 Case Selection Criteria

The assessment of this causal explanation will be conducted over Ukraine and Moldova as they are two relevant examples of non-direct EU mediation. In Ukraine’s case, EU has been a direct mediator for a brief period under the Geneva format in April 2014 (Elgström et al. 2018:305), however, unsuccessful attempts at de-escalating the conflict lead to its demise quickly. Currently, the peace talks are carried out under the Normandy format, led by France and Germany, for which the EU declared openly its full support (EEAS2015:2). In Moldova’s case, EU has taken upon an observer role in the 5+2 peace format talks with Transnistria since 2005 (Cristescu and Matveev 2011:11). While the negotiations collapsed in 2006, the format has been given new impetus since 2011, placing it again high up on EU’s conflict resolution agenda (Wolff 2012:6), especially after the Hamburg OSCE Ministerial meeting of 2016 (OSCE 2019). The selection criteria of these cases has therefore been theory-guided, as it is meant to test Bergmann and Niemann’s (2015:960) conditions for effectiveness which are intended ‘to be applicable to the full universe of cases of EU mediation’. Thus, based on this premise, the case selection criteria I followed is least-likely, by choosing two cases meant to test the confidence in a particular theory (Rohlfing 2012:84; Lamont 2015:133). This suggests that if EU’s values in all the independent variables have been high and resulted in an effective outcome in both types of EU’s mediation, then the theory ‘can make it anywhere’, whereas if it did not, then the reverse holds true (Bennett and Elman 2007:173; Rohlfing 2012:202).

3.3 Research Method and Conceptualisation of the Causal Mechanism

The method employed for assessing the causal explanation behind the effectiveness of EU’s different types of mediation will be that of causal process tracing (CPT). Generally speaking, CPT is employed as an analytical tool when the research goal is to extract a comprehensive overview over a temporal sequence of causal-processes part of a certain event or phenomena (Bennett and Elman 2007:183; Collier 2011:824; Blatter and Haverland 2012:81; Lamont 2015:135). More precisely, CPT is a technique best suited for questions ‘of which causal conditions are necessary or sufficient for the outcome of interest’, or for investigating theory-based ‘mechanisms’ that link causal factors to outcomes (Blatter and Haverland 2012:80). CPT is therefore deemed to be a within-case level research method, as it relies upon a configurational thinking where causal factors

(18)

18 become ‘necessary conditions’ and the resulting combination of factors leading to an outcome become a ‘sufficient condition’ (Rohlfing 2012:12; Blatter and Haverland 2012:82).

Regarding the causal factors of interest to this thesis, Bergmann and Niemann (2015:963-964) contend that their causal conditions are in fact interrelated, being treated differently only for purposes of analytical clarity; conditions such as choice of mediation strategy, for example, depending greatly on the availability of EU’s leverage. In a CPT approach, Beach and Pedersen (2013:109-110) contend that such causal theories need to be reconceptualised into causal mechanisms, in order to understand the ‘dynamic transmission of causal forces that produce an outcome’. Thus, based on the relationship among the necessary conditions described by Bergmann and Niemann, the causal mechanism of interest is best conceptualized as a ‘causal conjunction’ (see Figure 1.), which is a configuration where ‘multiple causal conditions work together (in additive or interactive ways) at a specific point or over a short period of time to produce the outcome of interest’ (Blatter and Haverland 2012:94).

(19)

19 3.3.1 Data Operationalisation

Having conceptualized the sort of causal mechanism this research operates on, a discussion on the operationalisation of each part of the conjunction is in order. Beach and Pedersen (2013:110) note in this regard that each part of a mechanism needs to be operationalised in a manner that ‘maximizes the certainty and uniqueness of the predicted observable implications’ while making sure that the evaluation is empirically feasible, meaning that data can be collected and measure what is intended to measure. Thus, by drawing from various available literature on EU’s peacebuilding practices, negotiation effectiveness, as well as its involvement in Ukraine and Moldova’s conflict resolution processes (Bercovitch et al. 1991; Tocci 2008; Levitsky and Way 2005; Sasse 2016; Beyer and Wolff 2016; EU Council 2009; Beardsley et al. 2006; Thomas 2012; Wolff 2012; Konstanyan and Meister 2016; Trilateral Contact Group 2015; OSCE 2019), table 2 offers below an overview of the empirical manifestations traced for evaluating the independent variables, hypotheses about the sort of findings needed to prove EU’s effectiveness and case-specific indicators used for measuring the dependent variable. Subsequently, the table is followed by a discussion on how each part of the causal mechanism was operationalized.

(20)

20 Necessary Conditions fostering EU mediation effectiveness (Independent Variable) EU-specific dimensions for observation Expectations for an effective outcome

Mediation outcome indicators (Dependent Variable)

Leverage:

Resources and instruments the EU can exert to spur an

agreement ‘the more resources

(leverage) the EU possesses, the more likely it will be an effective mediator’

(Bergmann and Niemann 2015:961-962)

Diversity of linkages (Sasse 2013:556-557)  Economic linkages (trade

and energy relations)  Aid linkages (financial

assistance or subsidies)  Social linkages (migration

trends and remittances)

Expectation1. The

more developed EU’s linkages towards the warring parties are the more likely the EU will be

a successful mediator

Expectation2. The

stronger Russia’s linkages are towards

the warring parties the more likely the EU will not be a successful mediator

Ukraine

Settlement of Major Conflict Issues (Minsk II):  Withdrawal of heavy

weapons by both sides  Immediate and

comprehensive ceasefire  Ensure effective OSCE

monitoring and verification of the ceasefire and weapons withdrawal

 Release and exchange all hostages and other illegally detained people  Safe delivery of

humanitarian aid to those in need

 Pardon and amnesty of figures involved in the conflict

 Restoration of full social and economic links with affected areas

 Full Ukrainian control over its border with Russia throughout the conflict zone

 Withdrawal of foreign armed groups from Ukrainian territory  Constitutional Reform regarding the decentralisation and special status acknowledgement of separatist regions

Settlement of Minor Conflict Issues (Minsk II):  Modalities for local

elections on special status of Donbas under

Ukrainian Law

Moldova

Settlement of Major Issues (package of eight):  Enhanced joint law

enforcement with the exchange of data on criminal cases and unimpeded travel of officials of the two sides  Freedom of movement

for people, goods and services, and  Apostilization of education qualifications issued in Transnistria  Acceptance of Transnistrian vehicle registration plates to the international traffic  Access for Moldovan

farmers to the

Transnistrian Dubasari region farmlands  Ensure functioning of

Moldovan-administered Latin Script Schools Settlement of Minor Issues

(other points specified in Berlin 2016 protocol):  Signature by both sides

of protocols regarding cooperation in the area of meteorology and environmental protection in the Nistru River Basin  Open the Gura

Bicului-Bicioc bridge  Reintegration of

telecommunication link between the conflicting sides

Strategy:

Strategies EU can adopt to foster an agreement between

conflicting sides (facilitation, formulation,

manipulation) ‘the more “manipulative” the EU’s mediation strategy, the higher the chances the mediation effort will lead to a conflict settlement’ (Bergmann and Niemann 2015:962)

EU types of mediation -

Concept (2009): 1. Direct mediation: EU

engaged in high-level mediation as a lead- or co-mediator, employing the full range of instruments available at its disposal to foster an agreement

2. Mediation promotion: EU

promotes dialogue and mediation non-coercively based on its own

experience as a peace project

3. Mediation leveraging: EU

provides diplomatic leverage (e.g. as part of a group of friends) and/or economic credence to mediation processes, and support follow-up to their outcome through the full range of civilian and military crisis

management instruments and policies targeting

Expectation 3. If the EU leverages mediation efforts or acts as a direct mediator in a conflict resolution

process, then the higher the chances

the mediation will contribute to a

conflict settlement/abatement

than if the EU were to only promote, support or fund the

(21)

21

trade, development and crisis response

4. Mediation support: EU

facilitates mediation processes led by other actors through capacity building, training, logistical support and the provision of expertise to mediators and conflicting parties

5. Mediation funding: EU

provides financial support to formal, informal and grassroots mediation processes

 Local elections in Donetsk and Luhansk according to OSCE standards

 Trilateral working groups between the Ukrainian government, the separatists and Russia Conflict Transformation

(Lanz et al. 2008)  Effectiveness and Impact:

Conflict situation (security situation, level of violence, humanitarian situation) improved as a result of mediation (confidence building measures, economic incentives or sanctions)  Coverage: Inclusion of

relevant issues, and regional and societal actors in the mediation process (neighbouring states, domestic civil society groups and other minority groups). Conflict Transformation (Lanz et al. 2008)  Effectiveness and Impact: Conflict situation (security situation, level of violence, humanitarian situation) improved as a result of mediation (confidence building measures, economic incentives or sanctions)  Coverage: Inclusion of

relevant issues, and regional and societal actors in the mediation process (neighbouring states, domestic civil society groups and other minority groups).

Coherence:

Degree of coordination and substantive agreement between Member States’ policies towards a conflict and EU’s mediation activites

‘a highly coherent EU

approach towards a particular conflict sends a strong signal of EU’s determination to help the warring parties to solve their conflict, leaving no room for interpretation concerning EU’s preferencesregarding the substance of a negotiated solution

(Bergmann and Niemann 2015:963)

Foreign policy behaviour EU Member States employ to promote the outcome intended by the Union’s common policy (Thomas 2012:459):

 threats and reassurances they communicate to others, where they commit their resources and what agreements they accept or reject

Relevant EU member states for mediation processes:

 Ukraine: Germany and

France

 Moldova: Germany and

Romania

Expectation 4. If

EU’s foreign policy and those of its Member States most

relevant to the conflict resolution

process are coordinated to suit

the common goal, then the higher the

chances EU will send a strong signal of commitment and impact positively the

settlement process

(22)

22 3.3.2 Explanation of Indicators and Data Evaluation

Condition I: Leverage

The causal condition leverage is operationalised in this causal mechanism as EU’s degree of linkage with the warring parties instead. This choice is motivated both by Bergmann and Niemann’s (2015:962) conceptualization of leverage, as well as their hypothesis for an effective outcome, which in their original form would obscure the empirical observations of this thesis’ CPT approach. That is because while Bergmann and Niemann’s conceptualisation of leverage builds directly on Tocci’s (2008:882-886) logic of EU’s positive and negative conditionality, it omits the structural factors strengthening or limiting EU’s potential exertion of leverage, such as conditionality’s objective (value of EU benefits mapped against the cost of compliance with its obligations) or subjective (relationship of the warring parties with the EU) value. Thus, the assumption they generate regarding EU’s effective exertion of leverage states that ‘the more resources (leverage) the EU possesses the more likely it will be an effective mediator’. Conceptualized in this manner, tracing EU’s leverage would entail an ex post assessment of the mediation process which would have to take into account the outcome variables in order to understand what the ‘more resources’ are, without providing additional information on the underlying action-formation mechanisms linking the causes to the effect.

Consequently, an alternative way for tracing EU’s degree of leverage without prejudgement to the causal mechanism and direction of causality, is to look at its degree of linkage towards the warring parties (Sasse 2013:555). Understood as the density of a country’s ties to an international power, linkage has been conceptualized alongside leverage (governments’ vulnerability to external pressure) by Levitsky and Way (2005:21-22) to explain the successes and failures of Western actors’ democracy promotion in authoritarian regimes. Both of these factors are expected to raise the costs of authoritarianism, however, leverage (manifested as politico-economic conditionality, diplomatic pressure or military intervention) has rarely had a sufficient democratizing effect on a regime on its own, without an extensive degree of Western linkage (socio-economic and geopolitical relations). Linkage has been considered to raise the costs of authoritarianism by increasing the odds of an international response to government abuses, by creating influential constituencies with a stake in adhering to democratic norms or strengthening democratic forces in relation to autocrats. Thus, linkage is considered to have the potential of turning into leverage,

(23)

23 being argued by Sasse (2016:291) to provide for a better starting point in process tracing EU’s amount of leverage, as it points out to sets of relationships which can be mapped more comprehensively at a certain moment in time.

In the context of unresolved post-Soviet conflicts, the concept of linkage can additionally account for the issue spoiler problems, included within Bergmann and Niemann’s (2015:963) discarded variable of conflict context, being already incorporated in a number of studies exploring EU and Russia’s competing interests in influencing the conflict dynamics in Ukraine or Moldova (Bayer and Wolff 2016; Hughes and Sasse 2016). Linkages can be explored from a variety of angles, Levitsky and Way (2005:22-33) differentiating between at least five different types (economic, geopolitical, social, communication and transnational civil society linkages). The general consensus is that the more developed linkages are the more likely they will contribute to the international powers’ degree of influence (leverage) (Sasse 2013:580; Beyer and Wolff 2016:348). However, since this research cannot feasibly account for all sorts of linkages EU and Russia have with the warring parties, the operationalisation of leverage is limited to three different types of linkages, identified by Sasse (2013:556-557) to have more prominence in the context of EU’s democracy promotion in its eastern neighbourhood. These types of linkages are: economic (trade and energy relations), aid (financial assistance) and social (migration trends and remittances).

Economic linkages include the trade and energy relations the parties have with the EU and Russia, as they provide for a better insight into the economic dependencies the parties have with the mediating actors, and how their diverging interests might influence the parties’ cost/benefit calculations in the mediation process. For example, energy relations are considered to be an important economic linkage next to trade, cutting across also aid linkages in Russia’s case through the preferential gas delivery prices acting as a form of external aid, Ukraine and Moldova’s energy dependencies on Russia being an important factor shaping their foreign policies towards Moscow, but also its influence in their domestic policies (Sasse 2013:557). Subsequently, aid linkages refer to EU and Russia’s financial assistance to the parties, considered also as in-built energy delivery subsides where relevant, providing a sense of the type of influence they can exert over the sides and their priorities within the peace process. However, in the case of unresolved conflicts, Russia’s aid linkages also include military assistance. Assessing the volume of this assistance is beyond the

(24)

24 scope of this research, nevertheless, it will be taken in consideration when assessing EU’s linkage with the warring sides, as Kramer (2008 cited in Sasse) points out to the fact that Russia’s military assistance provides a ‘much more direct linkage than Western aid’. Finally, social linkages look at the sides’ number of migrants to the EU versus Russia providing an idea of linkage in terms of people-to-people contacts, hence, geopolitical persuasion, while their remittances measured in % of the GDP provide a sense of their economy’s exposure to external factors (Sasse 2013:563).

Consequently, there are two expectations generated in relation to this causal condition, the first one states that: the more developed EU’s linkages towards the warring parties are the more likely the EU will be a successful mediator. By the same logic the second expectation asserts that: the stronger Russia’s linkages are towards the warring parties the more likely the EU will not be a successful mediator.

Condition II: Strategy

EU’s mediation strategy is operationalised according to the types of mediation described in the Concept document of 2009. In order to evaluate the type of mediation describing EU’s actions in each case, an assessment of its policies towards the negotiation process and implicated parts will be conducted with reference to its self-conceptualisation in the document. The hypotheses generated regarding the effectiveness of different mediation types, rests on the mediator strategies proposed by Bergmann and Niemann (2015:962), facilitation, formulation and manipulation, which contend in line with other literature (Bercovitch et al. 1991:16; Beardsley et al. 2006:67-70) that the more manipulative EU’s mediation strategy is, the more likely it is to lead to a conflict settlement, or abatement. An important mention stressed by some scholars in this instance is that these different types of mediation should only be seen as having different chances of success in situations where they are employed separately from one another, Wallensteen and Svensson (2014:320) suggesting that manipulation is often seen as most effective when is combined with facilitation and/or formulation. Consequently, in order to determine how EU’s mediation involvement fits a certain strategy, I will refer to the table proposed by Beardsley et al. (2006:66),

(25)

25 which describe the characteristic tactics employed by a mediator in each of three possible strategies (table 3).

Table 3. Beardsley et al. (2006:66) mediation styles and observable tactics

Therefore, based on these tactics and the description of EU’s types of mediation from its Concept (2009:6) document, the most effective types of mediation are expected to be direct and leveraging considering they interfere the most with the negotiation structure, whereas, promotion, funding and supporting are expected to be the least effective types of mediation, only if employed separately from the former two. Therefore, the expectation for this causal condition reads that: if the EU leverages mediation efforts or acts as a direct mediator in a conflict resolution process, then the higher the chances the mediation will contribute to a conflict settlement/abatement than if the EU were to only promote, support or fund the process

Condition III: Coherence

EU’s coherence followsthe degree of coordination and agreement between the Union and its individual Member States’ foreign policies towards a common conflict resolution goal (Bergmann

(26)

26 and Niemann 2015:963). Thus, the aspect Bergman and Niemann suggest to look at is EU’s political cohesion, which according to Thomas (2012:459) reflects how fully EU actors support whatever common policy has been agreed by the Union towards a particular conflict. Analysing political cohesion requires familiarity with the content of the common policy in place and detailed evidence of what the Member States do to support it, including the threats and reassurances they communicate to others, where they commit their resources and what agreements they accept or reject. Given the interrelational nature of the independent variables, the policies I will refer to when tracing Member States’ political cohesion will be those reflecting EU’s mediation strategy and usage of linkages. Considering the analysis cannot fully comprise all of EU’s 27 member states’ policies, the Member States I will be focusing on will be those whose involvement in the conflict resolution process is most relevant for the overall context. Thus, for Ukraine, EU’s degree of coherence will be analysed in reference to France and Germany, given their direct involvement in the conflict mediation process and their influence on EU policy (Janning and Möller 2016:4; Konstanyan and Meister 2016:2). In Moldova’s case, two relevant countries for the conflict resolution process are Germany, given its interest taken in solving the Transnistrian problem throughout the years as well as its special relation with Russia (Ivan 2014; Wolff 2012:23), and Romania which besides its ‘deep historical and cultural links’ with the country has expressed its commitment towards the conflict resolution process at the 2010 OSCE Astana Summit (Wolff 2012:23). However, some broader reflections upon EU’s general coherence will be conducted on certain issues of notoriety that stood out in the resolution process, attempting to convey a fuller picture of EU’s overall determination to help the warring parties. The expectation generated for this dimensions reads as follows: if EU’s foreign policy and those of its Member States most relevant to the conflict resolution process are coordinated to suit the common goal, then the higher the chances EU will send a strong signal of commitment and impact positively the settlement process.

Assessment of causal conditions

In order to stay consistent with CPT’s configurational nature, based on the generated expectations my analysis will estimate EU’s leverage, strategy and coherence as having either high, moderate or low chances for success. These values are merely operationalized to allow in the mediation outcome section for a clearer evaluation of ‘which causal conditions are necessary

(27)

27 or sufficient for the outcome of interest’, as well as observation of the causal mechanisms’ hypothesized effect over the mediation outcome.

Thus, based on Expectation 1-2, for EU’s linkages to be considered moderately effective, this would entail three possible situations, as shown in Figure 2: 1. EU’s economic, aid and social ties outweigh Russia’s in regards to only one warring party, 2. EU’s linkages outweigh Russia’s in regards to both parties but only in one or two dimensions, 3. EU shares an equal amount of linkages with Russia in at least one dimension but from both warring sides, for example when an economic linkage is characterized by EU-dominated trade and Russian-dominated energy relations. In a situation like this, EU’s amount of leverage, and chances of positively impacting the mediation process, will be considered moderate. Conversely, in any situation where EU/Russia’s linkages are higher or lower than in the configuration presented, EU’s leverage chances for mediation effectiveness will be accordingly considered low or high.

Regarding Expectation 3., if the EU combines a manipulative strategy, expected to have high effectiveness, with a facilitative one, expected to have low effectiveness, its overall strategy will be considered high, since the literature suggests that a combination of such strategies usually entrenches the value of the more effective one rather than deduct from it. Given that coherence works by a logic of subtraction, Expectation 4. will be evaluated contrary to Expectation 3. logic, where a medium degree of coherence on EU’s behalf can be the result of a Member State’s high degree of cohesion with the EU and another one’s low degree, or the result of a medium degree of coherence combined with a high degree.

(28)

28 Mediation Outcome

Lastly, the dependent variable mediation outcome has been operationalized according to the objectives stipulated in the peace protocols currently in place, Minsk II for Ukraine and the ‘package of eight’ for Moldova (Trilateral Contact Group 2015; OSCE 2017). Thus, for reasons of practicality the ceasefire and no agreement indicators have been eliminated, whereas the full agreement one is left implicit upon the fulfilment of all stipulated conditions. The separation of objectives into major and minor issues has been done deductively according to the information available regarding the importance of conditions (Konstanyan and Meister 2016; Bentzen 2018; OSCE 2018; de Waal 2018; Hill 2018). Subsequently, regarding the indicators for conflict transformation, these are based on the OECD Guidance on Evaluating Conflict Prevention and Peacebuilding Activities (Lanz et al. 2008) and will assess whether any similar outcomes occurred as a result of EU’s mediation and whether EU had an impact at conflict-transformation level. 3.3.3 Temporal Focus

The designated period of time this research will process trace will be 2014-2020. The reason for which I will be starting with 2014 is because the Ukrainian crisis and Russia’s annexation of Crimea brought renewed attention to the post-Soviet secessionist conflicts in EU’s Eastern Neighbourhood, especially in Moldova where the Transnistrian parliament requested to join the Russian Federation (Nilson and Silander 2015:53). Moreover, the consequent signature of the Association Agreements with both of these countries in 2014 changed EU’s standing towards these countries into a greater position of influence, where Moldova and Ukraine committed themselves to ‘enacting substantial legislative and policy change in line with EU’s demands’, especially on issues such as conflict prevention or crisis management (Delcour 2018:490-493). Moldova’s different conflict context is acknowledged in this thesis, however, the reason for which I will refer to the period of 2014-2020 is to allow for analytical proportionality as much as for feasibility. Therefore, while EU’s first intervention in Moldova happened in 2005, the negotiation format reached a halt in 2006, being resumed only in 2011, one year the after negotiations for the Association Agreement commenced (Government of Republic of Moldova, n.d.). Moreover, EU’s mediation involvement in the 5+2 format resumed only in 2014 through EUBAM, following Moldova and Ukraine’s accession in the DCFTA, and the concomitant hostilities in Ukraine (EUBAM 2015:25), which resulted in the first conflict resolution protocol being drafted only in

(29)

29 2016 after the 5+2 negotiations in Berlin (IPN 2016). This means that an earlier period to 2014 would not have allowed for an operationalisation of conflict outcome indicators at any rate. Instead, while it could be possible to start tracing the period of 2016-2020 for Moldova, the choice of 2014 is made in order to account for EU’s change of foreign policy and analyse its impact on the later signed protocols of 2016.

3.3.3 Generalisation and Validity

Given the double analytical focus of this thesis, explained earlier as lying at both cross-case and within-case level, the generalisation value of the ensuing results needs to be carefully reflected upon. First of all, CPT is a within-case technique of causal inference, which means that the combination of ‘necessary’ conditions, which together become a sufficient condition for an outcome, are strictly confined to the cases under study. Blatter and Haverland (2012:82) term therefore CPT’s generalisation value as “possibilistic”, meaning that the conclusions apply only towards the set of causal configurations that make specific kind of outcome (Y) possible. However, tracing a process that leads a causal factor to an outcome makes it possible to enhance the internal validity of a causal claim that ‘x matters’ (Gerring 2007 cited in Blatter and Haverland 2012:79). This sort of causal claim benefits from an ‘added value’ especially when the compared cases are not as similar as they should be, to be ‘controlled’, like when the measurement and classification of variables is not as clear-cut as it should (Blatter and Haverland 2012:79). Therefore, if this thesis’ analysis proves the EU-specific expectations, operationalized above, to be right, the results will implicitly enhance the internal validity Bergmann and Niemann’s causal claims, extending their generalisation value to the ‘entire universe of EU mediation’, since the case-selection fits the least-likely criteria.

3.4 Data Collection and Limitations

The primary data used for this research consists of official EU foreign policy documents such as financing decisions, council conclusions, protocols and press statements of conflict-resolution associated parties such as the OSCE or individual Member States. These include: ENI multi-annual programming documents, Member States’ National Governments factsheets, EC Special Implementing Measures and Joint Staff Working Documents, EEAS country relations factsheets or communications, EUBAM annual reports and OSCE press statements and documents. Lamont

(30)

30 (2015:82) notes however that in regards to document-based research the researcher does not, and cannot, have complete access to archives of an organisation or state, furthermore, should a document have complete information on a subject it should be noted that the data was inscribed in accordance to the institutional memory. Therefore, secondary resources such as academic articles, open-source journalism, think tank policy reports and factual data from other relevant institutions will be used to augment any potential knowledge gaps but also counterbalance the information contained within certain documents.

Some limitations pertaining to this study should also be considered. One main limitation pertains to the case study design of this research which Flyvbjerg (2006:234) notes that ‘allows more room for the researcher’s subjective and arbitrary judgement than other methods’. In endeavouring to study EU’s mediation involvement process this can be a very likely risk given the depth and breadth of EU’s conceptualisation of mediation, as noted by Haastrup (2015:3), which might render the selection and interpretation of certain data subject to the researcher’s own bias.

Another limitation pertains to the selected research method, Blatter and Haverland (2012:135-143) pointing attention to the fact that a CPT approach relies upon observations such as confessions about ‘perceptions, motivations and anticipations of major actors’ which are beyond the research capacity of this thesis given the structural limitations of a desktop research. Furthermore, while CPT is a within-case technique of analysis, it should be clear that it cannot account for the entire spectrum of causal processes behind an outcome, but rather focus on specific aspects. For that reason, the generalisation value previously mentioned only extends to the wider population of cases if analysed under the same configurations and data operationalisations.

Lastly, regarding the cases of Ukraine and Moldova, it should be stressed that information required for the CPT of causal mechanisms might differ in level of detail according to the condition or time period traced, the Ukrainian crisis being a more scrutinized topic during recent years than Moldova’s Transnistrian settlement. However, certain data might be equally difficult to collect despite the temporal difference in academic interest between the cases. Here, for example, Sasse (2013:557-563) points out to the lack of ‘good annual data on migration flows’ and impossibility to quantify Russia’s complex system of indirect subsidies to its post-Soviet neighbours, when process tracing Ukraine and Moldova social and aid linkages with their mediators. Therefore, the analysis of each part of the causal conjunction will strive towards providing an estimation of the

(31)

31 EU’s conditions for effectiveness rather than exact measurements of their value for the period of six years.

(32)

32

Chapter 4: Leverage

4.1 EU’s building of linkages and exertion of leverage: Ukraine 2014-present

4.1.1 Economic linkages (2014-present)

Ukraine’s overall economic linkage with Russia began to slide on a downward trend ever since 2013, when Moscow began to mount a series of punitive trade measures against it to deter its signature of the AA/DCFTA with the EU. However, the failure to sign the AA acted as a catalyst for the Euromaidan protesters, who saw European integration as the country’s ‘best and only pathway for modernization and economic growth’ (Wolczuk 2017; Euromaidan Press 2019). Therefore, after Yanukovich’s ousting in February 2014, Russia attempted to apply additional economic pressure on Ukraine by imposing another series of trade sanctions, which were subsequently intensified between April and August 2014 when Ukraine elected a new pro-European president and finally signed the AA/DCFTA (Cenușă et al. 2014:1). In the aftermath of these sanctions, as well as Russia’s annexation of Crimea and its support for the rebel militias in the east, Ukraine was left in a dire socio-economic situation with an annual budget deficit forecast at 7% of the GDP in 2014 and by further 12% in 2015 (World Bank 2015a). In response to this situation, the EU temporarily eliminated all of its custom duties for Ukrainian exports, seeking to support the country’s economic stabilisation but also entrench its economic linkage in the vacuum left by Russia, pressuring Ukraine into restructuring its economy to adapt to EU standards and expand beyond the post-Soviet markets which it was so largely dependent upon until 2014 (European Council 2014a; Molchanov 2016:524-525).

According to World Bank data (cited in Bhutia 2019) Ukraine’s trade turnover with the EU in 2014 amounted to $38 billion while its turnover with Russia to $22 billion, being a significant drop from 2013 when it amounted to $38 billion. In % of the total exports and imports, EU had a share of 40% of Ukraine’s imports that year and a 31.8% of its exports, while Russia’s shares stood at 25% and 19% respectively (Wilson 2015; Ministry of Foreign Affairs Ukraine 2019). The gap significantly exacerbated in 2016 when Ukraine was unilaterally removed from the CIS by Russia, upon its entry in EU’s DCFTA. Ukraine’s total trade turnover with Russia during this year amounted to only $8billion while the one with the EU stood at $30 billion. In % of the total, EU’s

The different shades of EU mediation: A comparative analysis of EU's conflict resolution support for Ukraine and Moldova's unresolved secessionist conflicts (2014-present)