A plurivocal formula useful for political purposes

It is striking that the EU’s Political and Security Committee policy document explaining the scope and meaning of the formula of ‘non-recognition and engagement’ has never been made public – neither at the time of its endorsement, in 2009, nor any time later. ² In principle, it is in the EU’s interest to be transparent about its objectives and the instruments for attaining them. In this particular context, however, it was feared that disclosing such a document would lead to unnecessary polemics in Georgia, Abkhazia or South Ossetia, or even within the EU. ³ Since then, in official documents regarding Georgia’s conflicts, the EU has therefore referred to a ‘policy of non-recognition and engagement’ without elucidating its precise meaning and the means of its implementation.

From the very start, the EU formula was aimed at legitimizing a policy combining non-recognition with engagement in Georgia’s conflict. The EU needed to be present on territories that were not under the control of Tbilisi, in order to be better informed about the local security situation and for the sake of conflict transformation. The projects implemented within the framework of the EU’s engagement policy were aimed at bringing the positions of the conflicting parties closer together, in line with the EU’s role as mediator in Georgia’s conflicts after the 2008 war. The formula of ‘non-recognition and engagement’ reaffirmed the EU’s stake in the preservation of Georgia’s territorial integrity and its opposition to Russia’s policies in the region – Moscow having recognized Abkhazia and South Ossetia as independent states in 2008.

The formula is not part of an overall EU strategy for enabling the Georgian government to recover the two territories, but it does more than expressing a simple practice. It is actually to be considered a policy, in the sense that any contradiction between the position of non-recognition and the practice of engagement has to be decided in favour of the position on non-recognition. The EU also describes the formula as a ‘comprehensive approach’ (Council of the European Union 2018), which means that, in implementing it, it takes a wide range of factors into consideration. Georgia has been receptive to this position. In 2008, it promulgated a law on occupied territories which included severe legal restrictions on any form of foreign presence on the territories of Abkhazia and South Ossetia, but it was not fundamentally opposed to a limited form of engagement with their populations – in fact, in July 2010 it promulgated its own ‘Action Plan for Engagement’ (Office of the State Ministry for Reintegration and Civic Equality 2010).

The formula allowed the EU and Georgia to express their joint opposition to the recognition of Abkhazia and South Ossetia and their shared willingness to engage with the populations of these territories. And it allowed the EU to define its own priorities for engagement, in line with its assessment of its potential and risks, but also with the policy instruments and financial means at its disposal. Peter Semneby (2010) – who, as the EU’s special representative for the South Caucasus, was the main architect of this policy – explained in an interview in May 2010 that, for the EU,

both the non-recognition and engagement are indispensable parts of one policy, because non-recognition without engagement is counterproductive; it will only lead to raising barriers between the Georgians and people in these regions. But engagement without the non-recognition part is a policy that would mean a risk that any step that we take could be instrumentalized in various ways and misused to stake claims in this dispute about status. Engagement is not about status, but in order to disassociate engagement from the status we have to be clear where we stand on the status issue.

Academics, think tanks and NGOs dealing with the Georgian–Abkhaz conflict were eager to use the ‘engagement without recognition’ formula for their own purposes. Researchers did not necessarily want to express any normative position regarding the status issue. Most of them were therefore keen to turn an EU position on non-recognition into a simple description of the state of affairs, even though some of the authors and organizations using this formula still defended a prescriptive position on non-recognition. NGOs focusing on conflict transformation and resolution do not have to choose sides, and are generally agnostic on the issue of recognition or non-recognition – defending either of these positions would unnecessarily make their activities in conflict areas more difficult. Furthermore, their financial dependence on EU funds, and the need to rely on authorization from local authorities in order to operate in the breakaway regions, would also make it difficult to engage in an open debate on recognition versus non-recognition. The revised version of the formula was helpful in allowing them to call for a strengthening of the EU’s engagement policies – involving additional support for their own activities – without having to refer to non-recognition as a normative principle. The London-based NGOs Conciliation Resources and Safer World called on the EU to favour an engagement policy with more immediate benefits for the populations of the disputed territories: ‘EU engagement needs to be more tangible to ordinary people and better understood in order to gain maximum benefit from the policy of engagement without recognition.’ ⁴

The Abkhazian criticism of the lack of EU engagement on their territory should also be mentioned in this context. Those Abkhazian political leaders and NGOs that were striving for closer links with the EU turned the original formula of ‘non-recognition and engagement’ into an additional variant of ‘non-recognition and non-engagement’. ⁵ In their dialogue with the EU, this allowed them to be critical of the latter’s perceived lack of coherence: the EU’s position on non-recognition was seen as contrary to the objectives of confidence building and post-conflict reconstruction.

After reaching its primary objective – convincing the Georgian authorities that engagement policies did not constitute a threat to the EU’s policy on non-recognition – the formula in a revised form was thus open to interpretation by the various actors, thereby enabling the EU to hold an open dialogue with all the conflicting parties and other stakeholders. The expression was allowed to mean different things to different audiences: each individual actor was free to choose one of the formulas and invest it with their own preferred substance.

It may then be asked why, in documents and bilateral agreements, the EU has generally confined the official use of such a successful formula as ‘non-recognition and engagement’ to Georgia’s conflicts, even though it defends a policy of non-recognition and the need to engage in many other conflicts too. This is even more remarkable as it uses policy instruments similar to those it has adopted for Georgia in other conflicts in its Eastern Neighbourhood, for instance in the conflict in Moldova on Transdniestria. The EU’s Eastern Partnership policy, launched in 2009, adds new opportunities for engagement to the existing instruments of the European Neighbourhood Policy. These new opportunities are open to both Georgia and Moldova. They include association agreements (AAs) and deep and comprehensive free trade agreements (DCFTAs), each offering new policies on establishing links with the populations living in disputed territories, including on visa liberalization and new financial instruments designed to support conflict transformation (Fischer 2013). The use of the same policy instruments for engagement in both conflict settings did not, however, give the EU sufficient reason to extend its use of the formula of ‘non-recognition and engagement’ to Transdniestria. Nor has it been used in other conflicts on the international status of territories where the EU is involved.

The formula was not extended to Transdniestria or to other breakaway territories for several reasons. First, unlike in Georgia, there was no pressing need to convince the Moldovan or any other central government confronting secession that the EU’s policy of engagement with a breakaway territory was not leading to shift in its position on non-recognition. Second, neither the Council nor the Commission had an interest in encroaching on the reserved powers of the EU member states regarding the question of recognition and non-recognition. The introduction of such a general formula could raise complex political and legal issues, and the ensuing discussion would then be divisive. Third, any attempt to extend the validity of a policy formula to other forms of engagement that are constrained by a policy of non-recognition would – even in a highly abstract form – go against states’ existing practice of not making their preferences regarding such a policy explicit. The lack of a formalized, systematic set of precepts reflects the need for an ad hoc approach to these questions. The ‘non-recognition and engagement’ formula therefore remained limited as a formal description of the EU’s policy towards Georgia’s conflicts, even though EU officials were free to use it in a personal capacity in workshops and informal meetings when describing other policies and practices, relating for example to Transdniestria.

A plurivocal formula in need of analytical precision

The academic literature on the EU’s policies of engagement has extended the geographical scope of the ‘engagement without recognition’ formula far beyond Abkhazia, South Ossetia and Transdniestria. It has been used to study the EU’s policy in its Eastern Neighbourhood (Harzl 2018; Newman and Visoka 2018) and Cyprus (Kyris 2018), but has also been seen as applicable to all forms of interaction between the EU, or even the US, and non-recognized entities that are located in Europe, Africa or Asia (Ker-Lindsay 2015; Berg and Pegg 2016). It has been regarded as capable of creating a common systematic framework for comparing all such cases.

This choice is problematic. A comparative analysis that brings together all forms of EU engagement policies in non-recognized states under the single formula of ‘engagement without recognition’ overlooks the necessary distinction between two types of EU engagement policy. The fact is, a position of non-recognition means more than the absence of recognition, and more even than the statement that the unrecognized territories cannot count on future recognition by the EU. Regarding Northern Cyprus, Abkhazia, South Ossetia, Transdniestria, Donetsk, Lugansk, Taiwan and Somaliland – entities which all EU members refuse to recognize as independent states – the EU plays a coordinating role by reinforcing a policy of non-recognition when its institutions and member states are engaging with the population of these territories. Here, the formula of ‘non-recognition and engagement’ is applicable. In Kosovo and Palestine, by contrast, where its member states are divided on the issue of recognition, the EU engages in line with a status-neutral position. And, here, its coordination policies can be described as ‘engagement without recognition’. ⁶

Policies of non-recognition and policies of engagement have different objectives. Non-recognition is generally chosen out of a concern to support long-term objectives relating to the stability of the international order, such as respect for existing borders. The objectives to be reached through engagement are generally more concrete, and often shorter-term. They include the implementation of humanitarian programmes, mediation and confidence building between conflicting parties, or the establishment of strong trade links between them or with the EU. It is assumed that attaining these short-term objectives will, step by step, allow for more trust between the parties, and that it might even be able gradually to transform the conflict setting in a positive way.

In a policy of non-recognition and engagement, the refusal to recognize an entity as an independent state comes first, and the pursuance of particular goals regarding this conflict, to be achieved through engagement, second. A policy of non-recognition delimits the acceptable forms of engagement in each of the conflicts under consideration. The main goal to be achieved with the help of engagement is the strengthening of the means to reach the objectives of a non-recognition policy. The highest preference goes to engagement programmes that directly reflect those objectives, such as the creation of common security mechanisms for the conflicting parties, or a trade regime that creates mutual trust and interdependence. Some initiatives even foreshadow reunification.

A policy of engagement without recognition, by contrast, is based on the necessity for the EU to remain status-neutral. Here, the EU member states are divided on the objectives to be reached and/or the means for achieving them through policies of recognition and non-recognition, in particular where conflict resolution and regional stability are concerned. A status-neutral policy involves different kinds of constraints on engagement from a policy of non-recognition (Berkes 2017). Engagement here entails an obligation to search for pragmatic solutions and agreements between the conflicting parties that do not strengthen or weaken any status claims.

The two formulas thus reflect different types of coordination of engagement policies through the EU, each of which entails particular problems to be addressed. In the first type of coordination, all member states refuse to recognize a particular entity as an independent state, and here the EU reflects a consensual position. But each member state also has the discretionary power to give a particular interpretation to its position on non-recognition, and to turn it into a particular policy of engagement. It will then defend its position within the EU. Moreover, each of the EU institutions, in its own sphere of competence, has specific functions to perform regarding non-recognition and engagement, and will link these to a particular interpretation of these policies. This means that, here, the EU – the EU institutions and the member states – not only have to reflect a consensual position on the question of recognition but, furthermore, must also define forms of engagement that are acceptable to all member states and that reflect a minimum of coherence among the institutions.

Coordinating a policy of non-recognition and engagement

A number of issues to do with engagement in the defence of a non-recognition policy are divisive for the EU’s member states and institutions, such as the question of whether or not the territories concerned are occupied by a foreign power. This issue directly determines what is acceptable as engagement (Coppieters 2018). Under international law, occupation cannot lead to unilateral changes in the status of a territory. Annexation or an attempt to create a new state under such circumstances are to be considered breaches of law. The description of the territory of a self-proclaimed independent state as being occupied by an outside power thus turns a policy of non-recognition into a principle of non-recognition, and pragmatic considerations regarding the use of a policy of non-recognition to achieve international stability then give way to the legal principle of non-recognition. Non-recognition then ceases to be a discretionary decision for the member states, and the EU itself is under an obligation to resist the redrawing of international borders.

The EU’s member states and institutions are divided on whether or not territories such as Abkhazia and South Ossetia should be considered occupied. This means that their consensual affirmation of a position of non-recognition does not imply agreement on its precise meaning or on its significance for engagement. Here, the EU has to take on a coordinating role and defend a policy that all its member states find acceptable. Moreover, a consensus between EU member states regarding the condemnation of a breach of jus cogens norms – as in the case of the creation of a self-proclaimed independent Republic of Crimea and its subsequent annexation by Russia – may go hand in hand with differences of opinion as to how a defence of the principle of non-recognition should be turned into concrete policies. The EU has to take such disagreements into account when imposing sanctions on an occupying power or adopting policies of engagement with the populations of the disputed territories.

Each EU institution will have a different approach to bridging these differences. It is the European Council and the Council of the European Union that coordinate EU non-recognition and engagement policies. In some cases, this may require a consensual approach from the member states and in others a qualified majority – each requiring particular compromises between the member states. Moreover, the EU’s mediation efforts in the conflict regions, made by the Council and the European External Action Service (EEAS), require a prudential attitude. Even where EU diplomats are defending a principled position on non-recognition, they would be well advised to avoid terminology that is unnecessarily offensive to one of the conflicting parties or external actors. They also need to avoid any position that would limit their flexibility in defending a policy (or the principle) of non-recognition in diplomatic negotiations, or that would place additional constraints on engagement. For instance, they will not avoid using the terms ‘occupation’ or ‘annexation’ in the context of Crimea, but they will in the context of Eastern Ukraine. In the international negotiations with Russia on Eastern Ukraine, the use of such a term by the EU would certainly jeopardize the chances of successful conflict management within the framework of the Minsk agreements – as these international agreements are not based on the assumption that the regions of Donetsk and Lugansk are militarily occupied by Russia. The Council and the EEAS also avoid defending such a position in the context of Abkhazia and South Ossetia, as that would likewise make both the dialogue with the ‘de facto’ authorities and international negotiations with Russia on the future of these territories more difficult.

It is the European Commission (EC), meanwhile, that actually implements the engagement policies. The EC’s activities are extremely wide-ranging in scope. It and Taiwan are involved in multilateral trade negotiations within the framework of the World Trade Organization, and they also have intensive bilateral trade relations – Taiwan being the EU’s fourth-largest trading partner in Asia in 2016 (European Economic and Trade Office in Taiwan 2016). The EU’s engagement with Transdniestria, through trade relations, is also well developed. Generally, in territories under the control of non-recognized authorities, the EU gives funding for programmes in the health sector, infrastructure, education and support for NGOs. The EC is, moreover, very active in delivering humanitarian help for the benefit of local populations, as in the South Caucasus, for instance. Each of these activities requires an acceptable balance between the defence of its position on non-recognition and the need for the EU to be present in these territories. The European Parliament (EP), by contrast, is not hampered by such diplomatic constraints. From its institutional perspective, it is far more important to formulate principled positions for the general public than ones that allow for diplomatic flexibility, as is the case with the Council, the EEAS and the EC.

Other EU institutions – such as the European Court of Auditors (ECA) and the Court of Justice of the European Union (CJEU) – likewise have particular approaches to turning a position on non-recognition into a concrete policy. The ECA is only marginally involved in questions regarding non-recognition, but – when examining whether engagement projects are managed effectively and whether they have reached their objectives, for instance – it does defend criteria such as sustainability. These criteria are not easy to meet under conditions of non-recognition. In a 2012 report on EU engagement in Northern Cyprus, for instance, the ECA pointed to the lack of cooperation with the local authorities in Northern Cyprus as a major cause of the inefficiency of EU engagement projects (Coppieters 2018). The CJEU examines few cases relating to non-recognition and engagement. But when it does, as for instance in the case on external trade and property issues affecting Northern Cyprus (Talmon 2001), its role in creating a legal framework for EU engagement policies is considered crucial.

Coordinating a policy of engagement without recognition

When the EU implements ‘engagement without recognition’ policies, the coordination procedures are different from those in cases of non-recognition, and the Council and the Commission have to refrain from making any pronouncements regarding the status of a contested entity. In the case of Palestine, however, the division between EU member states concerns only its present status: there is a consensus among them in favour of a future two-state solution to the conflict. But when it comes to Kosovo, the EU can make no such distinction between time perspectives, which means that while it may very well make public pronouncements in favour of the future statehood of Palestine, it must remain neutral in its statements concerning the present and future status of Kosovo.

In such situations, the EP, by contrast, is not bound to a position of neutrality as regards status. In 2010 it adopted a resolution calling for all EU members to recognize Kosovo (Euractiv.Com 2010). It also managed, in 2014, to bridge the gap between a refusal to recognize the statehood of Palestine at present and the need to recognize it at some point in the future. By an overwhelming majority of 498 votes to 88, it voted in favour of a resolution stating the following: ‘the European parliament supports in principle recognition of Palestinian statehood and the two-state solution, and believes these should go hand in hand with the development of peace talks, which should be advanced’ (Beaumont 2014). The CJEU has a role in defining the legal limits of the EU’s engagement with Israel and with the Palestinian National Authority. It has, for instance, dealt with EU trade policy towards the territories occupied by Israel, and has also addressed the question of whether Hamas, which controls the Gaza Strip – a territory where the EU strives for active engagement – is to be considered a terrorist organization or not.

Comparing the two EU coordination policies

Governments confronting breakaway draw red lines for the EU’s presence in the breakaway territories, its relations with the de facto authorities controlling breakaway territories, and its position on the presence of these de facto authorities in multilateral frameworks. The prohibitions vary considerably from one conflict to another. Azerbaijan does not allow any EU presence on the territory of Nagorno-Karabakh or of its surrounding regions. Georgia watches very carefully to ensure that EU activities in its breakaway territories do not contradict its own policies. The People’s Republic of China (PRC) puts no restrictions on EU economic and cultural activities in Taiwan – as long as they are considered ‘non-political’.

The counter-secession policies of governments confronting secession have a strong impact on the EU’s engagement policies when they share a position of non-recognition. The EU then perceives the drawing of such red lines by these states as legitimate in principle – and even more so when the EU defends the legal principle of non-recognition. Then it will generally restrict its engagement policies to what the central government considers compatible with its counter-secession policies. This does not rule out friction, however, or even serious disagreements between EU institutions, on how to draw particular red lines. In the case of Cyprus, for example, in 2010 the Commission attempted to develop direct EU trade with the North, an initiative that was opposed by Cyprus and was not taken up by the EP either. Regarding Georgia’s conflicts, the EU Special Representative has criticized the restrictions contained in the 2008 Georgian law on occupation (Coppieters 2018).

By contrast, the EU is less dependent on the policies of a central government confronting breakaway when it follows a policy of engagement without recognition. EU status-neutral policies are defined on the basis not of the red lines set by counter-secession policies but of the need to avoid conflicts with member states on the subject of recognition and non-recognition. Neutrality on status does not permit the condemnation of unilateral moves by the unrecognized authorities to strengthen their claims to independence (Berkes 2017). But an EU policy of ‘non-recognition and engagement’ treats such forms of conflict escalation differently from an EU policy of ‘engagement without recognition’. The EU has repeatedly expressed its concern that the policies of Taiwan’s president Chen Shui-bian are ‘not helpful to maintaining stability and peaceful development in the Taiwan Straits’ (Lim and Winkler 2012: 184). This was particularly the case for his 2004 initiative to organize a referendum on UN accession under the name of Taiwan. The EU has also condemned the organization of referenda for independence in the breakaway territories of South Ossetia, Abkhazia and Transdniestria. It does not, on the other hand, formally condemn unilateral moves by Palestinian or Kosovar authorities to strengthen their claims to statehood and independence.

An EU policy of non-recognition does not necessarily support the status proposals of a government confronting breakaway, but it always presupposes that the final status of a disputed territory will be mutually agreed between the conflicting parties. The EU defends a policy of non-recognition regarding Taiwan, but this does not mean that it shares Beijing’s interpretation of the One China principle and its application to Taiwan – rather, the EU refers explicitly to ‘its own’ One China policy. It takes an open-ended position on Taiwan’s final status, including on such questions as sovereignty, international personality and citizenship, as long as this status is accepted by both Beijing and Taipei. The EU would also support an agreement between Azerbaijan and Armenia that would not provide for the reintegration of Nagorno-Karabakh into Azerbaijan. But in other cases, where it likewise defends a policy of non-recognition, the EU comes forward with a defence of single-state solutions, as for instance regarding Transdniestria, Abkhazia and South Ossetia. The OSCE’s 57 participating states have endorsed a status settlement for Transdniestria that is based on Moldova’s sovereignty and territorial integrity and allows for ‘a special status’ for Transdniestria within Moldova (Scanlan 2018). In the case of Cyprus, the EU’s defence of a single-state solution is prescribed by UNSC resolutions. They stipulate that a settlement ‘must be based on a State of Cyprus with a single sovereignty and international personality and a single citizenship, with its independence and territorial integrity safeguarded’ (United Nations Security Council 1992). All these examples of EU positions on status, which are based on a policy of non-recognition and engagement, refer to solutions that will eventually have to be agreed by the conflicting parties. By contrast, an EU policy of engagement without recognition is, precisely, based on a divergence between member states on whether the statehood of a self-proclaimed independent entity should be recognized before or after the parties have reached an agreement on this issue.

The distinction between the two formulas relates, furthermore, to whether statebuilding efforts may be part of a policy of engagement or not. It is a general rule that a policy of engagement in line with a policy of non-recognition does not allow for such endeavours. Reinforcing the effectiveness of the political institutions of the non-recognized entities, and the legitimacy of the self-proclaimed independent authorities, would run counter to the declared objectives of the policy of non-recognition. Engagement programmes aimed at strengthening institutions that exercise the traditional functions of a state – in the fields of maintaining public order, defence, economic management and the exercise of legislative or executive power, for instance – are therefore excluded. For the same reason, the EU’s policy of ‘non-recognition and engagement’ excludes support for democratic state reform, limiting its support to the democratization efforts of civil society institutions. Non-recognition rules out intergovernmental meetings and cooperation, and allows contact or cooperation between EU diplomats and members of non-recognized governments only on an administrative level.

In a situation where its policy is one of non-recognition, the EU will not use official titles such as ‘president’ or ‘minister’ to refer to self-proclaimed authorities. The addition of the term ‘de facto’ before such titles is, however, not always a good option, as these authorities consider it offensive. A policy of non-recognition must, moreover, avoid any form of engagement that might strengthen of the international legal personality of the non-recognized entity. This restriction may create tension between a policy of engagement and a policy of non-recognition. Consequently, it may also create tension within the EU: in 2010 the risk of strengthening the international legal personality of a non-recognized entity was key to the argumentation by the EP’s legal department against endorsing an EC proposal to facilitate direct trade of goods between the European Union and Northern Cyprus. The EP’s legal experts – unlike their colleagues in the EC – considered that applying the procedural rules that the Commission proposed to follow in establishing such trade links was tantamount to assuming that Northern Cyprus could be considered a ‘third country’ – an assumption that contradicted the principle of non-recognition (Karambelas 2010).

The EU supports a policy of ‘meaningful participation’ in international organizations by Taiwan – a formula it does not use in other conflicts on sovereignty where it likewise defends a policy of non-recognition. This is due precisely to its fear that the ‘meaningful participation’ formula could, in those other cases, be perceived as an attempt to strengthen the legitimacy and legal personality of a non-recognized entity. But the EU’s policy on Taiwan’s ‘meaningful’ participation in multilateral organizations reflects the particular attitude adopted by the PRC: Beijing accepts such participation provided statehood is not a condition for it, and provided Taiwan’s belonging to China is expressed in one way or another – as it is in the World Trade Organization (WTO), to which it belongs as a customs territory, under the name ‘Separate Customs Territory of Taiwan, Penghu, Kinmen and Matsu (Chinese Taipei)’. The EU’s position on Taiwan’s meaningful participation is based on the acceptance of these two conditions.

The non-recognition of statehood implies the non-recognition of the constitution of a self-proclaimed sovereign state. This then implies, in turn, the non-recognition of elections that are held according to constitutional rules. But the EU’s position on non-recognition allows for different policies regarding the democratic character of elections, as can be illustrated by the cases of Taiwan and Abkhazia. For the Taiwanese presidential elections of 16 June 2016, the EU supported the democratic process taking place on the island. This was in line with its position that it shares common values with Taiwan regarding democracy and the rule of law. It supported the elections without, however, mentioning the name of the candidate who won them, Tsai Ing-Wen: a clear reference to the name of the president-elect had to be avoided, as it could have been considered a form of recognition of a head of state. This means that the EU draws a distinction between the expressive character of elections – expressing democratic values shared with the EU – and their function of selecting representatives for state institutions. The first dimension can be recognized by the EU, but not the second. Federica Mogherini, the EU’s High Representative for Foreign Affairs and Security Policy, made a statement on the evening of the 2016 elections that stressed the first dimension, and also mentioned the need for stability across the Taiwan Strait: ‘today, the people of Taiwan have voted. Respect for democracy, the rule of law and human rights underpin the successful conduct of these elections. The EU reiterates its support for the continued peaceful development of cross-Strait relations’ (European Union External Action Service 2016b).

By contrast, the EEAS condemned the holding of parliamentary elections in Abkhazia in 2017. It stated that the EU ‘does not recognise the constitutional and legal framework within which these elections have taken place’. This position is also valid for the elections in Taiwan, as it is for all elections taking place in political entities that the EU does not recognize as states – but here it is made explicit regarding Abkhazia, whereas in the context of the elections in Taiwan the EEAS avoided referring formally to this rule. As with the Taiwan elections, the EEAS did not mention what kind of elections had taken place (referring merely to elections, in general), but – in contrast to its support for the democratic values reflected by the elections in Taiwan – it denigrated the Abkhazian claim that its elections could be considered an expression of democratic values, referring in its statement to ‘so-called elections’ (Delegation of the European Union to the International Organisations in Vienna 2017).

The EU’s policy towards Somaliland contains three exceptions to the policy of non-recognition and engagement. First, it increases the country’s domestic capacity in terms of public order and statebuilding efforts, for instance by giving military aid to Somaliland’s coast guard (European Union External Action Service 2015). Second, it directly supports the democratization of state structures. Somaliland’s ‘international partners’ (which include, apart from the EU itself, EU member states Belgium, Denmark, Finland, France, Germany, Ireland, Italy, the Netherlands, Sweden and the United Kingdom) actively supported the 2017 presidential elections. They noted, in a statement – which was also signed by the United States, Norway and Switzerland – that the observers ‘praised the smooth and peaceful conduct of voting and, despite areas of concern, concluded that irregularities were not on a scale such that they would undermine the integrity of the electoral process’ (European Union External Action Service 2017). Third, the EU does not avoid the use of official titles in referring to Somaliland’s leaders. The EU was even officially represented at the inauguration ceremony of President Muse Bihi Abdi in 2017 (European Parliament 2018).

To justify making such exceptions to its policy of non-recognition, the EU could argue that a policy of engagement that was not active in the fields of maritime security or democracy building would overlook the difficult regional context in the Horn of Africa, which would contradict the non-recognition policy’s long-term objectives of stability and respect for the international order. Moreover, the EU also has its own security interests to defend in supporting Somaliland’s military – for example in the fight against piracy. It combines this statebuilding policy with active support for the negotiations between Somaliland and Somalia, in line with its non-recognition policy.

Statebuilding, on the other hand, is the general rule for status-neutral policies, even if such efforts are not necessarily made under this heading. The EU aims to reinforce the key functions of statehood of Kosovo and the Palestinian Authority. In neither case does it have any issue with recognizing the legitimacy of the electoral process (including its function within the state institution) or with calling Kosovar or Palestinian leaders or authorities by their official titles. Kosovar and Serbian representatives take part in EU-mediated negotiations on an equal footing, and the documents refer explicitly to ‘Serbia’ and ‘Kosovo’. But at the same time the EU also refers, quite prudently, to the ‘relations between Pristina and Belgrade’ (European Union External Action Service 2016a).

The distinction between the two types of coordination is important for an assessment of the limits of EU engagement, but it does not give any indication of the intensity of the EU’s relations with entities whose international status is contested. These relations are not necessarily less intense under a policy of non-recognition than under one of being status-neutral: it is entirely possible for the EU to engage profoundly with a non-recognized entity without contradicting its own non-recognition policy. Since 2011, for instance, it has a visa-free regime with Taiwan, which is not the case for Kosovo or for Palestine. It considers Taiwan a fully fledged democracy, which is not the case either for Kosovo or Palestine. Its trade relations with Taiwan are more important than those with the other two. But, unlike with Kosovo or the Palestinian Authority, these intense, positive relations do not allow for meetings between high-ranking EU officials and their Taiwanese peers, or allow the EU to refer to them by their official titles: that would run counter to the EU’s policy of non-recognition in the context of cross-Strait relations.

The distinction between the two kinds of EU engagement is crucial for a better understanding of its capacity to strengthen the effectiveness of non-recognized entities. The extent to which these entities are made dependent on EU engagement, however, must also be taken into account. Kosovo’s and Palestine’s dependence on EU support is greater than that of Northern Cyprus (which depends mainly on support from Turkey), Nagorno-Karabakh (which depends solely on support from Armenia), Taiwan (which for its military security depends entirely on the United States), or South Ossetia, Abkhazia, Transdniestria, Lugansk, Donetsk or Crimea (which are all completely dependent on Russia). Up to now, Somaliland has survived without any patron state – which does not preclude increasing dependence on EU aid in the future.

The search for international and regional stability is an important factor in the EU’s policies on non-recognition, in its defence of the legal principle of non-recognition, in its neutral position regarding the question of status and also in its recognition policies. On the one hand, it is one of the main considerations when choosing between these four different options. The strengthening of state resilience, the defence of territorial borders and the containment of state fragility, on the other hand, take on a different meaning within each of these four policy frameworks, but also in terms of the actual security environment. The EU can thus support state effectiveness for Kosovo, Palestine and Somaliland despite its status-neutral policy regarding Kosovo and Palestine or even its formal denial of the statehood of Somaliland. Such support does not necessarily increase these territories’ chances of future recognition by all EU member states: state effectiveness is indeed one of the conditions of recognition, but the EU states may for instance consider that objectives such as international stability or respect for the international legal order are better served through a policy of non-recognition than through the establishment of full diplomatic relations.

Finally, the distinction between the two types of EU engagement policy coordination is important in order to analyse the kind of failure EU engagement may be confronted with. In case of a policy of non-recognition and engagement, engagement programmes may fail to be implemented, to become sustainable, or to reach standard objectives such as conflict transformation. The overall policy of non-recognition and engagement will, however, never be considered a failure by the EU itself so long as the latter can demonstrate that it was able to remain unyielding in defending its position on non-recognition. The failure of an engagement policy based on a status-neutral position, by contrast, cannot be justified in this way. Here, the EU has to demonstrate that it is able to overcome the divergent positions of its members on the question of recognition and non-recognition and to develop engagement policies that are successful. In Kosovo and Palestine, non-recognition does not provide for a fallback position if the EU’s engagement policies should fail.

Conclusions

The lack of clarification of the precise meaning of the EU’s ‘policy of non-recognition and engagement’ towards Abkhazia and South Ossetia allowed the phrase to be transformed by other actors into a general ‘policy of engagement without recognition’, and is one of the key factors in the wide acceptance of both formulas. There is a need, however, for an analytical distinction between a ‘policy of non-recognition and engagement’ and a ‘policy of engagement without recognition’. This distinction highlights the specific restrictive conditions for EU engagement. The first formula, in contrast to the second, refers to an EU engagement policy that excludes programmes aiming at statebuilding. This contrast is not without exceptions, however. In Somaliland, where stabilization corresponds to the long-term objectives of a policy of non-recognition, the EU is engaged in statebuilding. But, as in all cases of non-recognition, including Somaliland, its policy of engagement remains within the framework of a non-recognition policy and related objectives. It is hardly possible for the EU to support the participation of non-recognized entities in multilateral frameworks – its support for Taiwan’s ‘meaningful participation’ being a significant exception in this regard – but it does so when engaging with Kosovo or Palestine on a status-neutral basis. But in this context, the exception is also fully in line with the EU’s non-recognition policies: its support for Taiwan’s international involvement does not transgress the red line set by the PRC, which allows for such participation as long as it does not challenge the One China principle.

The EU is far more sensitive to the red lines drawn by counter-secession policies under a policy of non-recognition than under status-neutral policies, but then too the EU’s non-recognition policy is never identical to that of a central government confronting a breakaway – including the counter-secession policies of EU member state Cyprus. Relations with entities that the EU does not recognize on the basis of a common position of non-recognition among the EU member states are not necessarily less productive, or intense, than its relations with those that are not recognized on the basis of a status-neutral position. This is illustrated by EU–Taiwan trade relations and by the EU’s visa-free regime for Taiwanese citizens. Because, as indicated in the first formula, it is more of a priority for the EU to achieve the objectives of a non-recognition policy than those of an engagement policy, a failure regarding the latter will not challenge the legitimacy of its policies overall – as long as the former objectives are not defeated. According to the second formula, status-neutral policies represent a bigger challenge to the EU. A failure regarding engagement here would directly challenge the EU’s legitimacy, also taking into account the fact that its status-neutral policy is already the result of an internal division between its member states.