EU sectoral integration in the Eastern Neighbourhood: the case of Frontex-Moldova relations in border management

ABSTRACT

The article explores EU sectoral integration in the Eastern Neighbourhood. It shows that extant approaches to explaining EU meso-level engagement with third countries – differentiated integration and functional cooperation – fall short of capturing all dynamics at play. Drawing on the literature of ‘Europeanisation,’ it develops a novel conceptual framework for the study of EU sectoral integration entailing four avenues – rules, institutions, practices and knowledge – and two features – local ownership and cross-fertilisation. Through an in-depth empirical study of Frontex-Moldova relations, the article demonstrates that engagement in border management remains below the threshold of differentiated integration by precluding institutional inclusion but reaches beyond functional cooperation through combining integration concerning practices and knowledge. Furthermore, unlike differentiated integration, it allows for local ownership, and contrary to functional cooperation, it enables cross-fertilisation, even if these features are absent from the macro-level cooperation context. Thus, the article contributes to the refinement of the theoretical framework of European integration.

KEYWORDS:

• EU
• Frontex
• European neighbourhood policy
• Moldova
• European integration
• border management

Introduction

Over the years, the external engagement of the European Union (EU) has intensified across the globe. Its international cooperation programmes and network of diplomatic missions now span all continents, with its immediate neighbourhood enjoying a particularly high priority in this endeavour.¹ Since the early days, this process has been closely followed and documented by scholars. While some studies concentrated on ontology with the aim to define what kind of actor the Union is in the international arena (see Whitman, 2013), others assessed its actorness by looking at what it does and with what results (see Schimmelfennig, 2015). However, most works within the latter strand have focused on macro-level phenomena with sectoral engagement having sparked comparably little interest. Consequently, empirical evidence on the EU’s meso-level, external relations is still scarce (see Muhhina, 2022), and relations between third countries and sectoral EU bodies such as decentralised agencies remain even more under the radar (see Hofmann et al., 2019).

This is surprising given that Brussels has long prioritised technical cooperation in sectoral capacity building over political engagement in its neighbourhood through the involvement of its ‘panoply’ of sectoral instruments (Korosteleva, 2016). Such intentions were clearly highlighted i.e., in the 2015 review of the European Neighbourhood Policy (ENP) which assigned a prominent role to EU agencies in this process (European Commission, 2015a, 2015b).² The relatively few empirical studies reflecting this development generally understand sectoral engagement in the ENP space in terms of a loose form of ‘functional cooperation’ aiming at the unidirectional externalisation of EU preferences (see Freyburg et al., 2011) within a macro-level context where Brussels only pays lip service to ‘local ownership’ (Petrova & Delcour, 2020). This type of collaboration is seen to differ from ‘differentiated integration’ targeting ‘Western European’ states which are provided institutional access to EU sectoral arrangements such as the Schengen Area after their complete approximation to related EU standards (see Schimmelfennig et al., 2015).

Through presenting original empirical evidence on the cooperation of the European Border and Coast Guard Agency (Frontex) – an EU decentralised agency – with the border management authorities of Moldova³ – an ENP country –, the research contributes to the scarce attempts at a systematic classification of EU sectoral integration, particularly in the Eastern Neighbourhood. By addressing the simple question, ‘what are the characteristics of the Frontex-Moldova engagement?’ it demonstrates that meso-level relations between the Union and third parties may reach beyond functional cooperation by entailing a shared epistemic community and the pooling of resources, while at the same time remain below the threshold of differentiated integration in that relations preclude organisational inclusion. Described as ‘operational integration’ (OI), this type of engagement is shown to follow patterns of cross-fertilisation and local ownership even if these are absent from the macro-level cooperation context of the studied policy field. As such, the aim of the research is two-fold: to contribute to the theorising of EU sectoral integration by drawing attention to a missing link, and to offer an in-depth empirical study of EU sectoral integration that remains scarce in the literature.

The article proceeds as follows. First, it provides a brief overview of the extant conceptualisation of EU external engagement at the sectoral level. It then develops a novel typology for the classification of EU sectoral integration, and subsequently presents the research design of the article. This is followed by a short summary of the cooperation between the EU and Moldova in the border management sector, highlighting the role assigned to Frontex in this process. The fourth part delivers new empirical evidence on the Frontex-facilitated cooperation between the border management authorities of EU member states, Schengen-associated countries (SAC), and Moldova.

EU sectoral integration: a short conceptual overview

It is widely held that sectoral engagement between the EU and third countries is marked by different dynamics than relations at the macro level. Meso-level transactions are largely decoupled from ‘foreign policy prerogatives’ and instead follow ‘functionalist dynamics’ which reflect ‘sectoral patterns of interdependence’ (Lavenex, 2015), and the prevalence of professional goals over political considerations (Majone, 1997). At the same time, while the EU is often depicted as a closed entity, its borders have proven ‘fuzzy’ with its surrounding regions forming an ‘intermediate space’ between its internal territory and the external world (Christiansen et al., 2000). Accordingly, sectoral engagement between the Union and third countries has generally been described either in terms of ‘differentiated integration’ (see Schimmelfennig et al., 2015) or ‘functional cooperation’ (see Freyburg et al., 2009, 2011, 2015).

Differentiated integration is defined as the organisational integration of third countries into specific EU policies with related EU preferences becoming legally valid in these states through complete approximation. Examples include the Schengen Area or the EU’s internal market with the institutional participation of the non-EU members Norway, Iceland, Switzerland and Liechtenstein (Schimmelfennig & Winzen, 2019). Functional cooperation on the other hand is understood as a loose collaboration between the EU and third-country sectoral bodies (Freyburg, 2011). Directed primarily at the externalisation of EU preferences, the most common example for it is the ENP instrument which is seen to have been implemented via the embedment of Neighbourhood states into EU-led ‘transgovernmental networks’ (Freyburg et al. 2015) along the lines of what is often rather problematically described as ‘Europeanisation’ (Schimmelfennig, 2015).

Accordingly, a typology provided by Lavenex (2011) distinguishes between three circles of ‘EUropean integration.’ The first circle involves ‘full organisational inclusion’ whereby countries participate in the Union’s central legislative structures which is currently only possible for EU member states. The second circle – which mirrors differentiated integration – refers to a looser form of organisational inclusion. Here, third countries do not partake in EU-level legislation as they are excluded from central Union bodies, but instead gain membership in sectoral EU bodies, ranging from full membership with limited voting rights to participation with observer status. The third circle – reflecting functional cooperation – entails ‘technocratic’ cooperation in the context of transgovernmental networks.

Accordingly, empirical studies typically depict the engagement between EU sectoral bodes such as decentralised agencies and third countries either in terms of some form of differentiated integration (Asderaki & Markozani, 2019; Turhan & Yildiz, 2022), or as functional cooperation (see Rimkuté & Shyrokykh, 2019; Sagrera, 2014; Shyrokykh & Rimkuté, 2019). However, as argued in the current article, none of these approaches proved sufficient to grasp every facet of sectoral engagement, thus leaving room for further explanations. In part, this has to do with the fact that very few works have ventured to theorise EU sectoral integration in a systematic manner, and even less aimed to deliver in-depth empirical evidence.

As a result, we witness a stretching of definitions to the extent that, i.e., differentiated integration is now described as ‘any modality of integration or cooperation that allows states (members and non-members) and sub-state entities to work together in non-homogeneous, flexible ways’ (Lavenex & Križić, 2019, p. 3; see also Klose et al. 2023; Lavenex & Lutz, 2023). On the other hand, accounts such as that by Lavenex focusing specifically on third-country participation in EU agencies through the lens of differentiated integration speculate that further, ‘understudied forms of external flexible integration’ are possible (2015, p. 839). It is therefore worth exploring alternative ways of EU sectoral engagement and providing a more nuanced conceptual framework for use.

The four circles of EU sectoral integration: a novel conceptual framework

Building on the above, the current article first provides a novel typology of EU sectoral integration so as to give more nuance to its study. Naturally, all following classifications are to be treated as heuristic categories intended at grasping and organising intellectually what is in reality a complex and fluid phenomenon. Importantly, the research applies a principal-agent approach (see Kassim & Menon, 2003) and understands the EU as a ‘multilevel’ administration (see Egeberg 2006). From this it follows that EU sectoral bodies – such as decentralised agencies – are treated as inseparable elements of the EU’s institutional structure, operating externally as its proxies (see Egeberg et al. 2015) instead of functioning as actors ‘in their own right’ (Groenleer, 2014; Pollak & Slominski, 2009; Wonka & Rittberger, 2011). Consequently, the engagement of EU decentralised agencies with third countries is seen here as an integral part of EU external actorness.

The fourth circle: integration via shared knowledge and practices

Drawing on the theory of ‘Europeanisation’, the article offers an update to Lavenex’ (2011) above ‘three circles’ typology. In short, ‘Europeanisation’ is defined as ‘a process involving: (a) construction, (b) diffusion and (c) institutionalisation of formal and informal rules, procedures, policy paradigms, styles, ways of doing things and shared beliefs and norms which are first defined and consolidated in the EU policy process and then incorporated in the logic of domestic (national and subnational) discourse, political structures and public choices’ (Radaelli, 2006: 59). Consolidating this definition, the research identifies and defines four elements (rules, institutions, knowledge and practices) that are to serve as avenues for meso-level integration between the EU and third countries.

First, ‘formal rules’ – or simply, rules – ‘comprise of regulation […] in specific policy areas, […] political, administrative, and judicial processes, and […] the setup and competences of state and sub-state organisations’ (Schimmelfennig & Sedelmeier, 2005, p. 7). Second, ‘institutions’ are defined as ‘integrated systems of rules that structure social interactions’ (Hodgson, 2015, p. 501) and refer here to ‘organisation’ described as a type of institution which is ‘made up of groups of individuals bound together by some common purpose’ (North, 1994, p. 361). Third, ‘informal rules’, ‘procedures’ and ‘ways of doing things’ may all be captured in terms of practices that are defined as ‘socially meaningful patterns of action, which, in being performed more or less competently, simultaneously embody, act out, and possibly reify background knowledge and discourse in and on the material world’. As such, practices are understood to refer to ‘structures (or agency), ideas (or matter), rationality (or practicality), and stability (or change)’ (Adler & Pouliot, 2011, p. 4). Fourth, ‘policy paradigms and style’ are best approached via tracing transformations in knowledge. Knowledge is what ‘enables [actors] to act in accordance with expectations, irrespective of whether they like the role or agree with it. The key is the agents knowing what is socially accepted in a given setting or community’ (Checkel, 2005, p. 804). Importantly, each of these elements is treated as a stage of integration that presupposes the stage preceding it. For instance, integration regards rules is not possible in the absence of integration into related institutional frameworks, and so on.

Building on this, the article contends that there is a fourth circle to be included in the framework proposed by Lavenex, which is termed here are operational integration (OI). In short, we observe OI if the EU – via its sectoral bodies – and third countries rely in their engagement – to perform functions towards the fulfilment of their mandates – on integration in terms of knowledge and practices, but not in terms of rules or institutions. Conversely, if integration along the lines of knowledge and practices is complemented by integration in an institutional sense, we observe differentiated integration instead of OI. Then again, we observe functional cooperation instead of OI if engagement regards knowledge and/or practices between EU sectoral bodies and third-country authorities remains under the threshold of integration while neither institutional nor regulatory inclusion take place. More concretely, OI entails sectoral cooperation between the EU and third countries based on shared knowledge via the construction of an (i) epistemic community, as well as on shared practices involving (ii) the pooling of resources, but without regulatory or organisational inclusion.

According to Haas, an epistemic community consists ‘of professionals with recognised expertise and competence in a particular domain,’ and is built on shared criteria for the validity of knowledge, as well as on a shared set of normative and principled beliefs – incl. regarding problems to be addressed – that provide rationale for social action and a basis for common practices (1992, p. 3). Naturally, members of an epistemic community may not always understand everything the same way but have instead a shared knowledge base which informs their activities within every domain of joint practice. In other words, we observe an epistemic community in the context of EU sectoral relations if the activities of EU professionals – of sectoral bodies or member states – and third-country staff draw on a comprehensive, shared set of codified standards (see Skinner, 2013). As such, harmonisation reduced only to selected domains of joint practice between parties does not constitute an epistemic community.

Following Vestlund (2017), the pooling of resources has two components. First, it entails the sharing of various assets such as data, other types of know-how and material resources between engaging parties which they then rely on in their functions towards the fulfilment of their mandates. Second, it includes ‘a routinized division of labour, with participants mutually adapting and specializing in different types of tasks and different fields of expertise’ (Vestlund, 2017, p. 63). Resources thus pooled entail those applied in EU-facilitated joint activities and possessed by sectoral EU bodies and/or EU member states, as well as by participating third-party authorities. It is argued here that for sectoral engagement to pass as a form of integration, both components of resource pooling must be present in a ‘de jure’ manner. In terms of the first component, this refers to legally established routes for the sharing of data, know-how, and material resources. Regards the second component, this involves the de jure integration of EU officials into the chains of command of third-country sectoral authorities, integration the other way around, or the de jure establishment of new, joint chains of command between the parties.

As opposed to OI, differentiated integration additionally entails organisational inclusion via the employment of third-country nationals as staff members by EU sectoral bodies, and their participation in the management boards of these organisations with limited to no voting rights. At the same time, functional cooperation falls short of combining a shared epistemic community and de jure resource pooling. Finally, and as discussed above, integration along the lines of rules refers to regulatory inclusion involving parties with full voting rights in the regulation of a sector. This entails taking part in the drafting of, and the decision making over the founding regulations of EU sectoral bodies operating in a given policy field, as well as full voting rights in the management boards of these organisations. Due to this being regulated by EU central bodies, integration regards rules is currently only possible for EU member states (Table 1).

Table 1. Avenues and circles of EU sectoral integration.

		Avenues of integration
		Knowledge	Practices	Institutions	Rules
Circles of integration	‘EU membership’	✓	✓	✓	✓
	Differentiated integration	✓	✓	✓	X
	Operational integration (OI)	✓	✓	X	X
	Functional cooperation	✓/X	X	X	X

Potential features: integration along the lines of local ownership and cross-fertilisation

Furthermore, a distinction can be made between the above circles based on two potential features. First, unlike differentiated integration which, by default, necessitates complete approximation to EU standards in particular sectors, OI and functional cooperation may allow for the preservation of local ownership at the sectoral level by third countries. Drawing on communitarian approaches, local ownership is seen to involve the predominance of local actors in the shaping and implementation of shared initiatives following domestic preferences. As such, it is domestic actors that define problems, develop related policies and put these into action (Petrova & Delcour, 2020).

In more practical terms, the prevalence of local ownership has different implications for integration regards knowledge and practices. In the case of the former, we can expect to observe ‘cherry picking’ whereby third-country authorities download some EU preferences while refraining from downloading others, according to their own needs (see Bolkvadze, 2016). This is then to result in an ‘interplay of differential normalities’ whereby sectoral EU preferences co-exist with others in domestic use (Merheim-Eyre, 2017). As for practice-related integration, the prevalence of local ownership is expected to involve the EU operating primarily as a ‘facilitator of transgovernmental networks’ rather than a supranational entity (see Egeberg & Trondal, 2016, 2017). In other words, the EU is to mainly act – via its sectoral bodies – as a facilitator of engagement between member state regulatory bodies, third-country authorities, and other actors in a way that this engagement may continue beyond its involvement (Wolff et al. 2022, p. 1623). In other words, the EU may operate as a ‘vanishing mediator’ likely to ‘vanish in its intervention’ (Balibar, 2003, p. 331).

Second, unlike functional cooperation which is generally seen to advance the one-way externalisation of EU preferences (see Lavenex, 2004), OI and differentiated integration potentially allow for cross-fertilisation (see Eberlein & Newman, 2008, p. 29; Gstöhl & Frommelt, 2023), whereby mutual influence between EU bodies and member states as well as third countries is enabled (see Zeitlin, 2015). Thus, in this context, EU sectoral bodies are only to function as hubs for the promotion of EU-wide solutions (Levi-Faur, 2011; Majone, 1995; Thatcher & Coen, 2008), rather than central regulators (see Dehousse, 1997).

In more practical terms, the theory of ‘Europeanisation’ distinguishes between three ways in which preferences are transferred, namely via downloading (from the EU to the domestic level), uploading (in reverse) and cross-loading (from the domestic to the domestic level) (Aggestam & Bicchi, 2019; Major, 2005). Applied to the above framework, knowledge-related integration may thus not only involve downloading, but also uploading whereby third parties transplant their preferences into the shared set of codified standards of their epistemic community. In a similar manner, integration in terms of practices may entail cross-loading in the context of which engaging parties share with and take on from each other pooled data and know-how. Both uploading and cross-loading may take place via the contestation of existing preferences or through the construction of new ones (see Dandashly & Noutcheva, 2022; Table 2).

Table 2. Features and circles of EU sectoral integration.

		Features of sectoral integration
		Local ownership	Cross-fertilisation
Circles of integration	Operational integration (OI)	✓	✓
	Differentiated integration	X	✓
	Functional cooperation	✓	X

Research design

The data presented in the article was obtained via the combination of three qualitative research methods. First, it draws on a systematic document analysis aimed at collecting and processing all publicly available official plans of and reports on relevant activities, a balanced selection of which is referenced in the research. Both the collection and the selection were done with the involvement of officials from studied domestic authorities and Frontex, as well as local and international officials working independently of these bodies. Second, information obtained the above way was discussed in the context of 10 semi-structured interviews conducted with domestic, EU and international staff. Third, the research involved fieldwork visits to the headquarters of the Moldovan Border Police and to the international airport in Chisinau, where officials provided insights into their work in October 2019.

The data was collected for the period between January 1st, 2011 (the official start of cooperation between Frontex and Moldova) and June 23rd, 2022 (the day Moldova was granted EU candidate status) in the context of a purposeful case study design aimed at identifying information rich cases for the collection of empirical evidence (see Palinkas et al., 2015). Indeed, Frontex, an instrument in the EU’s ‘panoply’ is seen to possess an extensive external mandate (Busuioc et al., 2012). In addition, it operates in the border management sector,⁴ an area marked by high levels of cross-border interdependence which is considered a main driving factor of integration (Schimmelfennig et al., 2015). At the same time, it is the Eastern Neighbourhood, and particularly Moldova, that the EU has the strongest institutionalised ties with (Langbein, 2014).

Naturally, the selected case study is not without limitations. Chosen merely as an information rich case to apply the above analytical framework in empirical study, first, the generalisability of the findings across sectors, countries and regions remains to be evaluated by future comparative studies. Nevertheless, the case presented may serve as an example of what is possible in the context of EU external engagement through focusing on Moldova whilst it remained outside the EU’s accession framework. Second, causality including defining circumstances are not assessed. Third, potential constraints on sectoral integration, such as the presence of possible ‘veto players’ (Dimitrova & Dragneva, 2013) or the challenges Moldova’s disputed territories may pose are not considered.

Background: the EU-Moldova cooperation in border management and the role of Frontex

To apply the above framework in study, it is helpful to consider an illustrative case of EU external engagement at the sectoral level: the Frontex-facilitated cooperation between Moldovan and Schengen border management authorities. Before providing original empirical evidence in the following section, a short background of this engagement is presented below. This includes a summary of the macro-level cooperation between the EU and Moldova in the border management sector, followed by a discussion of the role assigned to Frontex in this context.

The macro-level cooperation context

The intensification of cross-border challenges such as illegal migration, terrorism and organised crime has prompted the EU to expand its cooperation both internally and with third countries in border management (Monar, 2013), leading to the traditional divide between the internal and external dimensions of EU security policy becoming ‘fuzzy’ (Niemann & Zaun, 2018). With the 1999 incorporation of the Schengen Agreement into EU law, the Schengen Area, a single jurisdiction in border management-related matters, became part of the EU’s structure.⁵ This resulted in differentiated integration between EU member states as well as Norway, Iceland, Switzerland and Liechtenstein, the four SAC (Schimmelfennig & Winzen, 2019).⁶ At the same time, the high level of interdependence between the Union and its neighbourhood prompted the Union to deepen its cooperation with further third states (see Gawrich et al., 2010), putting this matter high on the agenda in the ENP, and particularly under its Eastern Partnership dimension (see European Commission, 2004).

Accordingly, while the Union had initially prioritised trade-related approximation in its neighbourhood, this strategy was later revised with priority instead given to cooperation towards inducing administrative and judicial reforms in partner countries (Wolczuk et al., 2017). Moldova, along with Ukraine and Georgia, is regarded as a ‘front runner’ in undertaking such reforms, having harmonised large parts of its legislative framework with EU standards (Börzel & Lebanidze, 2015). In the field of border management, areas such as migration, asylum and cross-border crime prevention were most affected, as demonstrated by the comprehensive, EU-induced institutional transformations witnessed in these areas. This was largely driven by Moldova’s ambition to join the EU, resulting in its willingness to undertake costly reforms towards convergence with Union standards and participate in any of its initiatives made available (Baltag & Bosse, 2014, 2016).

The transformations took place in the context of two macro-level frameworks. First, the Association Agreement signed by Moldova and the EU in 2014 aimed at the establishment of close economic as well as political cooperation and mentioned border management as a key area.⁷ Importantly, while conditionality applied in certain policy areas covered by the Agreement, border management remained outside its scope. Listed as a key sector for cooperation, here, the EU could only formulate recommendations instead of requirements,⁸ leaving partner countries the freedom to converge with standards at their own discretion (Delcour, 2013). This mirrored the EU’s internal structure where security-related matters have primarily been handled in an intergovernmental, and, to a lesser extent, transgovernmental manner, while the Union has had close to no supranational capacities in this area (see Wolff et al., 2013).⁹

However, the second incentive came with considerable levels of conditionality both in legal terms and concerning knowledge-related harmonisation. Moldova was the first ENP country to be provided with a visa liberalisation arrangement by the Union, followed up by action plans detailing concrete transformations to be undertaken by Moldova in border management.¹⁰ Related requirements negotiated at the macro level by the European Commission and the Moldovan government included, i.e., legal convergence with EU standards in the areas of migration and asylum, or the consolidation of the legal framework for border management so that domestic institutions share powers and functions in a more EU-conform manner. Most importantly, it also entailed the adoption of action plans and domestic strategies on the implementation of the EU’s Integrated Border Management (IBM) standards in domestic training programmes, and related adjustments to national ethical codes (European Commission, 2011).

In short, IBM refers to a comprehensive, shared set of codified standards originally developed for the management of the external borders of the Schengen area so that systematic internal border checks could be eliminated. It contains practical know-how covering all areas of border management, and a normative dimension with emphasis on the principles of fundamental rights, anti-corruption, and good governance (EaP-IBM, 2015a, p. 4). In other words, by signing up to the visa liberalisation scheme, the Moldovan government agreed to a full convergence with EU border management standards towards knowledge-related integration.

In addition, cooperation between the two parties was aided by the involvement of the EU’s ‘panoply of instruments and actors,’ including its decentralised agencies (see Korosteleva, 2016). 2011 saw the establishment of ‘The Eastern Partnership Panel on Migration, Mobility and Integrated Border Management,’ a regional consultative process between EU member states and Eastern Neighbourhood countries. The consultations aimed at the advancement of the dialogue on migration and asylum policies, the exchange of best practices, and the strengthening of asylum and migration systems in Eastern Partnership countries.¹¹ The Union’s 2020 strategy on the Eastern Neighbourhood highlighted the importance of ‘increased cooperation with EU justice and home affairs agencies towards domestic security sector reforms, joint fight against trafficking in human beings and illicit drugs, and integrated border management to improve partner countries’ ability to withstand the pressures they face and step up their resilience’ (Council of the EU, 2020).

In a similar manner, a 2015 European Commission review of the ENP explicitly stated the EU’s willingness to ‘further intensify work on information exchange, administrative capacity and operational and technical cooperation […] via training, research, capacity-building projects and pilot projects with partner countries, notably working through relevant EU Agencies, including Frontex and Europol’ (European Commission, 2015a, p. 17). Simultaneously, Moldova showed demand for the intensification of cooperation with EU agencies (see Malmström, 2012). This resulted in a considerable broadening of the association of the country with agencies, and particularly with Frontex (European Commission, 2015b, p. 6).

The role of Frontex

Frontex was established in 2004 with tasks related to the management of the external borders of the EU and SAC. The founding regulation of the agency was amended in 2007 and 2011, then replaced in 2016 and in 2019. With each change, its external mandate got further extended and after the 2019 regulation, it covered the full spectrum of its remit initially set out for EU-internal operation only (European Union, 2019; Interviewee #9). Accordingly, Frontex has been described as an ‘operational-cooperation agency’, the mandate of which revolves around improved external cooperation (Busuioc, 2016). In addition, the external remit of Frontex was constructed in a way that it ensures the agency operating as an integral part of the EU in its external relations. More specifically, the Frontex regulation states that the agency shall always ‘act within the framework of the external action policy of the Union’ (Article 73:2). At the same time, its international strategy was constructed to form ‘an important pillar of the broader technical and operational strategy for the European IBM that Frontex is tasked to develop.’ (Frontex, 2017: Annex XIII).

The agency concluded working agreements with close to twenty third-party countries as well as regional, international and civil society organisations until mid-2022.¹² Its Working Arrangement with the Moldovan Border Police was signed in 2008 (Frontex, 2008), while the agency did not have a similar agreement in place with the Moldovan Customs Service in the studied period. The implementation of the Agreement was then facilitated through 3-year cooperation plans,¹³ via daily contacts between appointed officials on both sides, and in co-organised expert groups focused on specific issues.

The Agreement covered the full mandate of the agency except for the use of executive powers (i.e., investigative capabilities and coercive measures). In general terms, areas covered included joint operations, trainings, information exchange, risk analysis, the sharing of technology incl. information systems, as well as research and development. While these activities could involve Frontex staff, at times, the role of Frontex could also be reduced to the coordination of ‘activities for one or more member states and third countries [in] joint operations’ (European Union, 2019: Article 36c). More concretely, three competences of key importance were integrated into the remit of the agency over the years, through which Frontex got established as the principal sectoral organ of the EU with a mandate reaching into all areas of EU-wide shared competence in border management both within the Union and vis-à-vis third countries.

First, in 2011, Frontex was given primary responsibility for the facilitation of the IBM. The 2011 amendment to the Frontex regulation set out the promotion and implementation of the IBM as the main task of the agency both across the EU and externally.¹⁴ The regulation passed in 2019 then clearly stated that ‘the Agency shall provide technical and operational assistance to third countries’ in line with IBM standards (Article 71:2). As such, the agency gained ownership of the IBM, and with that, the responsibility to transfer this comprehensive set of codified standards to Moldova towards constructing a shared epistemic community in border management.

Second, the 2019 regulation also saw the integration of the European Border Surveillance System (Eurosur) into the agency’s mandate (Article 73:4). Eurosur was established in 2013 as a framework for information exchange between member states ‘to improve situational awareness and increase reaction capability at the external borders.’ It stores information gained via land, maritime and air border surveillance, as well as during checks at border crossing points and in border operations. With the help of Eurosur, Frontex is able to create up-to-date ‘situational pictures’ of movements along the EU’s borders and pre-frontier area so as to facilitate integrated planning and response. After 2019, access to the situational picture, initially provided only to Schengen states, could also be shared with third states.¹⁵ This made the potential pooling of resources in data in border management between the EU and Moldova possible.

Third, the regulations foresaw the possibility for agency-facilitated joint operations with third-country officials joining as observers, that is, the pooling of human resources between third countries, EU member states, SAC and Frontex in border management in a de facto manner. However, in 2022, the EU signed a status agreement with Moldova, enabling Frontex to use executive powers in the country through the deployment of ‘border management teams from the European Border and Coast Guard standing corps’ (European Union, 2022). The standing corps is the first uniformed service of the EU consisting in large part of member state officials and to a lesser extent, Frontex staff. Its primary task is to support Schengen states, while it can also be deployed to third countries.¹⁶ The possibility to deploy the standing corps with executive powers thus enables the de jure integration of EU member state, SAC and Frontex staff into Moldovan chains of command.

EU sectoral integration with Moldova in border management

Since the Schengen Agreement was made part of EU law, regulatory integration into the Schengen system has only been possible for EU member states. With Frontex having been installed as the principal sectoral organ of the EU’s panoply with a mandate that reaches into all areas of shared sectoral competence, the differentiated integration of SAC in border management entailed organisational inclusion into the agency through participation in its management board with limited voting rights, and by joining its staff.¹⁷ Apart from these four states, the cooperation of the EU with other third countries in border management was meant to remain below the threshold of differentiated integration. At the same time, as the below case study demonstrates, through the facilitation of Frontex, Moldova engaged in operational integration with the EU on the basis of integration concerning knowledge and practices.

Integration regards knowledge: the establishment of an epistemic community

The main idea behind all [training] activities was to create a bridge between EU border guards and their colleagues in the Neighbourhood so that we can have a common culture and a common understanding of the problems we encounter. (Interviewee #9)

Forming an epistemic community as geographically broad as possible in the field of border management had long been an aim of Frontex, particularly since it was tasked with the implementation of the IBM. The first ‘International Cooperation Strategy’ of the agency foresaw ‘an increasingly connected global border management community that lives up to the highest standards for border,’ which was to be achieved through ‘the advancement of harmonisation and enhanced interoperability’ (Frontex, 2017: Annex XIII). In line with this, Moldova drafted its first own National Strategy on Integrated Border Management as early as 2011 (EaP-IBM, 2015b), which was then followed by an even more ambitious domestic IBM strategy in 2015 (Guvernul Republicii Moldova, 2018).¹⁸ This process was actively supported by Frontex via the provision of trainings to Moldovan border police and customs officials (Frontex, 2020b; Frontex and MIA Moldova, 2021).

Between 2011 and 2014, Frontex was invited to Moldova to deliver trainings on standards in various areas covered by the agency’s mandate such as joint operations, risk analysis, document security and anti-trafficking in the context of an EU-funded project focused on the promotion of IBM in the Eastern Neighbourhood. Implemented by the International Centre for Migration Policy Development (ICMPD), the project trained 60 Moldovan border officers (Interviewee #4).¹⁹ In the same period, Frontex officials also supported the European Union Border Assistance Mission to Moldova and Ukraine (EUBAM) in its training activities which aimed at ‘helping embed the concept of IBM into domestic practice’.²⁰ These efforts were followed by the far more ambitious ‘Eastern Partnership Integrated Border Management Capacity Building Project’ (EaP-IBM) that ran between July 2014 and December 2018 under the management of Frontex.²¹ It is with this project that the agency became primarily responsible for IBM externally (Interviewee #9).²²

Importantly, the activities of this project focused on transplanting the Frontex Common Core Curriculum (CCC) for border police and customs education into domestic training systems.²³ The CCC is a comprehensive set of border management standards for education that seeks to encompass all state-of-the-art practices and normative principles seen as relevant for border management through the lenses of IBM. Thus, with trainings provided on the modules of the CCC, for the first time, Frontex engaged in the external promotion of know-how related to all areas of its competence, and by large, to that of the EU in border management. In other words, the CCC provided a foundation for a cross-border epistemic community in the sector. In practical terms, the transfer of the Curriculum included the promotion of EU-wide standards related i.e., to the detection of falsified documents, x-ray interpretation, the classification of goods, screening and identification of nationalities, interview techniques, anti-trafficking, technical English, cargo declaration, risk analysis, as well as human rights, transparency and good governance (Frontex et al., 2015, 2017, 2018a).

Although the project focused primarily on the externalisation of EU preferences, there was some room left for cross-fertilisation whereby Moldova could upload its preferences through shaping the content of the CCC. In 2017 alone, Frontex invited representatives from 13 non-EU countries, including Moldova, to take part in the updating of the common curriculum (Frontex, 2018). As a Frontex official stated, ‘[third-country] experts took part in working groups together with experts from [Schengen] states, and the updated Curriculum is a collective product’ (Interviewee #10). Another officer from the agency’s training unit added that ‘During our trainings, we had the opportunity to share our experience and show our colleagues [from Eastern Neighbourhood countries] how we deal with certain issues, while we could learn from their experience’ (Interviewee #9).

In addition, regardless of the macro-level EU-Moldovan agreement towards full convergence with the IBM, domestic authorities could secure a fair amount of local ownership at the meso level. As a project report attests, ‘the new [domestic curriculum] was elaborated at the national level based on local expertise’ (Frontex et al., 2018b). Indeed, all above trainings were delivered according to domestic needs identified by Moldovan authorities (Frontex et al., 2018a; Interviewee #3). In the case of the EaP-IBM project, this was done in a way that discussions were first held at the regional level, and then with relevant national authorities separately to cover domestic needs. At the same time, the project allowed for a revision of such needs in the context of annual work plans. Then again, the organisation of regular ‘project focal point meetings’ guaranteed that all activities were discussed in advance and planned out jointly (Civitta, 2016). As a Moldovan diplomat put it, while Frontex applied the same general framework in its cooperation with all participating countries, it did so in a ‘quite flexible way,’ leaving ample room for tailoring to local needs. As such, this process ‘worked in a way that Frontex told what it could offer, and Moldova took what it needed’ (Interviewee #1). The updated curriculum was then a domestic product to be approved and implemented by the Moldovan Ministry of Education which functioned independently to all other involved bodies (Frontex et al., 2018b).

In addition, Moldovan authorities exercised cherry picking. The CCC was not transplanted in its entirety but got diffused into different modules of the domestic curriculum. In addition, unlike in Schengen states, the level of implementation varied from chapter to chapter and subject to subject, depending on the extent to which topics were considered relevant locally. In more concrete terms, in Moldova, certain CCC modules were offered only as optional courses to students. The focus of such modules varied from ‘cross-border crime’ to ‘administration,’ ‘vehicle inspection and forensics,’ ‘tactical procedures incl. domination by weapon’ and ‘general knowledge on air, land and sea borders.’ However, every student had to take at least 15 classes in two of these modules. In the case of refugee rights, only 10 classes were compulsory but anyone in training could choose to do more. Other modules on ‘asylum and international protection procedures’ and ‘force-related situations,’ ‘European integration,’ ‘international legislation,’ ‘European code of police ethics,’ ‘Frontex code of conduct’ and ‘European border policing’ were not implemented due to their perceived irrelevance (Frontex et al., 2018b/ Interviewee #9).

Thus, Frontex acted as a hub for the promotion of EU-wide solutions rather than a central regulator, which resulted in a domestic curriculum in Moldova in which CCC standards co-existed with other standards following local preferences. However, the knowledge base of the Moldovan Border Police and the Customs Service was still found to draw on a set of codified standards shared with Schengen states in all domains of possible joint action to the extent that both authorities could be invited to participate in any form of activity covered by the agency’s mandate (see the below section). This was the outcome of Moldova having had ‘established a training system [based] on Integrated Border Management’²⁴ with most parts of the CCC getting successfully written into the domestic training curriculum.

While the implementation of some parts of the IBM officially started in Moldova in September 2013, it was with the help of the EaP-IBM project that local integration of the CCC really took off (Frontex et al., 2018b; Interviewee #4). The efficiency of this process was guaranteed through the use of the ‘train-the-trainers’ model whereby only a relatively small number of Moldovan officials needed to be trained who could then pass on their knowledge to colleagues (Frontex, 2018a; Interviewee #9).²⁵ Vocational trainings by local trainers were then delivered at the Border Police National College in Ungheni, with regular, 22 months courses offered to future sergeants, and at the ‘Stephen the Great’ Police Academy in Chisinau providing four-year academic trainings to future officers. The first cohort of the vocational training graduated in the summer of 2015, while that of the academic training in 2017 (Frontex et al., 2018b).

An assessment carried out locally in 2016 following a request by Frontex showed that the level of implementation of the CCC in Moldova was already higher at the time than in the accession countries Albania, Bosnia and Herzegovina, North Macedonia, and Montenegro. More specifically, in Moldova, 60% of the CCC was totally, while another 28% was partially implemented (Frontex, 2016). At the same time, an evaluation of CCC-related knowledge completed in 2017 with the involvement of 134 Moldovan officers in their final stage of studies and recent graduates came back with exceptionally good results. Answers provided by the officers had the accuracy of 70% regards ‘practical skills,’ 64% concerning ‘general studies’ and 61% in ‘law enforcement studies.’ Compared to the above accession countries, Moldovan officers performed better (Frontex et al., 2018b).

Trainings delivered by Frontex to trainers were accompanied by focused workshops offered to local staff employed in capacities which were key for joint action. This involved, i.e., the schooling of a handful of Moldovan officials in the ‘Common Integrated Risk Analysis Model’ (CIRAM) which also forms part of the CCC. Consequently, Moldovan border authorities promptly implemented the model seen useful for the identification of domestic priorities in risk analysis (Guvernul Republicii Moldova, 2015a). As a Moldovan border official highlighted, most domestic institutions in the security sector started to ‘speak the same language’ when it comes to risk analysis, while earlier, different authorities used different terminologies and approaches.

For instance, risk analysis used to mean something else to the each department and body of the Ministry of Internal Affairs. But recently, the Ministry approved [the CIRAM] which is now applied across all its units, and which was first applied by the Border Police, and then transmitted to other [national] institutions by its trained officers. (Interviewee #7)

Harmonisation based on IBM and the CCC was further aided by voluntary institutional transformations at the domestic level (Interviewee #5). First, the Border Police got integrated into the Ministry of Internal Affairs, thus transiting from a military-style organisation into a police-like institution that employs both armed officials and civil servants (Guvernul Republicii Moldova, 2012, 2018). Second, in 2011, the Moldovan government set up the National Council of Integrated Management of the State Border, consisting of representatives from various domestic authorities involved in the application of IBM. The council was tasked with the development, implementation and monitoring of the national IBM strategy, the coordination of all relevant activities, as well as the drafting of related policy recommendations, and legislative and regulatory frameworks, thus further cementing Moldovan local ownership over the IBM.²⁶ Third, 2014 saw the establishment of the Joint Risk Analysis Group with the aim to strengthen risk analysis capacities at the national level on the basis of the CIRAM (EaP-IBM, 2015b; Guvernul Republicii Moldova, 2015a, 2015b; Interviewee #6).²⁷ Finally, the National Coordination Centre of the Border Police, responsible for the coordination of operational activities with other authorities, was also subject to significant developments (Guvernul Republicii Moldova, 2015b), resulting in part in the setting up of mixed mobile teams that consisted of both border police and customs officials (Politia de Frontiera, 2015 Interviewee #6).²⁸

In sum, the Frontex-supported harmonisation in IBM-based training between EU member states, SAC and Moldova contributed significantly to the formation of a shared epistemic community between these parties. On Moldova’s end, this was further aided by organisational transformations in line with the IBM. As a European Commission official stated, ‘it became unnecessary to run any further training projects in [Moldova, and] only some focused, ad-hoc trainings may be provided in the future in the case of need’ (Interviewee #2). As such, after 2018, Frontex could largely ‘vanish in its mediation’ as a knowledge provider. At the same time, harmonisation left ample room for local ownership with Moldovan authorities exercising control over policy development and implementation incl. via cherry picking, while cross-fertilisation was also present through the uploading of Moldovan preferences into the shared Curriculum.

Integration in terms of practices: the pooling of resources

EU officials coming [to Moldova] without listening have largely been unsuccessful, while those that listened have in most cases been successful. (Interviewee #4)

The above knowledge-related harmonisation supported by Frontex and embarked on by Moldovan border management authorities with their Schengen counterparts increasingly aided practical cooperation between the two sides, resulting in a comprehensive pooling of resources. First, in the context of the EaP-IBM project, Frontex and Eastern Neighbourhood countries set up the Eastern Partnership Risk Analysis Network (EaP-RAN). Established in 2014, the network hosts 3–5 meetings yearly at the headquarters of Frontex with the participation of Schengen states and Eastern Neighbourhood countries, including Moldova. The meetings focus on challenges concerning risks at borders and irregular migration trends aiding the drafting of weekly and annual reports with statistical data. As of 2019, the parties have also held separate ‘Airport Border Risk Analysis’ network meetings twice annually with the attendance of border management officers and analyst experts deployed to airports in Schengen states and Eastern Neighbourhood countries. The meetings have focused on general migration trends, trafficking, and illegal migration at airports (Interviewee #9).

Both networks thus enabled the pooling of resources regards data with every country providing and receiving information. New data gained in the networks fed into the work of participating authorities, making them increasingly reliant on one another in an area widely seen to form ‘the starting point’ of all activities in border management from strategic decision making to the planning and implementation of operational activities.²⁹ The role of Frontex was to facilitate these networks and compile and analyse provided data on incidents, migration trends, patterns, threats and routes, as well as on enforcement actions, refusals and, overstays through its Risk Analysis Unit and with the help of network members (Frontex, 2022a; Interviewee #10).³⁰ The networks drew largely on the above-mentioned CIRAM methodology which thus provided a shared set of codified standards for risk analysis (Frontex, 2020b). However, in the case of the collection of certain data, participants could still draw on alternative methods which they had used domestically and saw fit. In some cases, elements of these methods – such as indicators – were taken up by other network participants (Interviewee #10), thus allowing for cross-fertilisation not only regarding data but also know-how.

Second, the exchange of data and know-how took place on a regular basis in other settings as well as at ad-hoc meetings and via Eurosur. An interesting example was the 2022 international conference that brought together law enforcement and border management officials from EU member states as well as Moldova and Ukraine ‘to discuss the changing situation at the borders’ considering the 2022 Russian invasion of Ukraine. While focused on a matter of key importance to the EU, it took place in Moldova and was co-organised by Moldovan and Frontex officials, enabling national ‘participants [to share] lessons learned from their respective operational responses to the war in Ukraine.’³¹ Furthermore, since the singing of the status agreement between the EU and Moldova, Moldovan authorities have had full access to Eurosur. Data shared via this system concerned illegal border crossings, the facilitation of illegal migration, the violation of the terms of stay, entry refusals, asylum seekers, document fraud and travel document forgery (Frontex, 2022a). Interestingly, the provision of access by the EU was a result of Moldova’s recommendation to reform the agency’s founding regulation. As a Moldovan diplomat put it, ‘We told them that the European legislation should be clearer so that cooperation [regarding Eurosur] could further proceed between us’ (Interviewee #1).

Finally, starting in 2011, Moldovan border police and customs officers participated in numerous Frontex-led joint operations, while Frontex and EU staff were deployed to Moldovan Coordination Points (CP), and later to Focal Points (FP), thus ensuring a pooling of human resources. The involvement of third countries in Frontex-led joint operations carried out by EU and SAC border authorities increased significantly over the years. In 2019 alone, 193 third-country observers participated in Frontex-led missions (Frontex, 2020b). Moldova stood out in this group by participating in an exceptionally high number of operations (Frontex, 2012, 2013a, 2013b, 2014, 2015).³² In these operations, Moldovan officers were deployed as ‘third-country observers,’ and thus integrated into joint chains of command in a de facto manner. However, the legal framework as to the roles observers could be given in joint operations was rather vague. Indeed, the boundaries between cooperation and observation have not been clearly set, with the latter being often defined along the same lines as the former (Marin, 2020). As a Moldovan border officer put it, the participation of his colleagues in Frontex-led operations went increasingly beyond simple observation as ‘mainly we are not observers, but we have made operational arrangements with Frontex’ (Interviewee #1).

As an example, operation Alexis that took place between April and June 2015 at the Henri Coandă airport in Bucharest, Romania, involved Moldovan border officials in the identification of risks and threats such as transits without visa or valid travel documents (Interviewee #8). Another example is the Joint Action Days Aeolos operation implemented in September 2017, and which led to 18 arrests at EU airports, with Moldovan officials having actively supported both the investigations and the arrests (Frontex, 2018). In addition, Moldovan officials also participated in Frontex-led return operations which were to assist EU member states in organising the returns of people residing illegally in the EU (European Union, 2011). These operations were based on ‘EU procedures on return, readmission and reintegration, [and] an integrated return management system in line with the best EU standards’ (Frontex, 2022b).

The role assumed by Frontex in all these operations was to mainly coordinate related activities carried out by national authorities. What is more, most of these operations were implemented primarily on Schengen territory, thus allowing for a blurring of geographical and functional borderlines between the EU, SAC and Moldova. At the same time, while joint operations ensured the de facto operational integration of Moldovan officials into the work of EU member state and SAC authorities on Schengen territory, Coordination Points and Focal Points established in Moldova enabled something very similar the opposite way. Here, however, the integration of EU and SAC officials into Moldovan chains of command happened de jure, in particular at FPs where these officials could use executive powers.

In short, CPs allowed for the deployment of the Frontex standing corps to Moldova with observer status. As a Frontex official summed up, the goal was to support host authorities with border checks, the detection of falsified documents, fraudulent identities and cross-border crime. Furthermore, CPs were to support the exchange of technical know-how as well as data on ongoing migratory trends (Frontex, 2019a, 2019b; Interviewee #10). Attested by a Frontex report on CPs, ‘EU Advisers/Observers deployed [to Moldova and other third countries] increased mutual understanding of many aspects of border control’ (Frontex, 2019a, 2019b). Between 2012 and mid-2022, Frontex set up four CPs in Moldova. Three were established along the border between Moldova and Ukraine – near Palanca–Maiaki-Udobnoe, Criva–Mamalya, as well as Otaci–Mohyliv-Podilskyj – and one at the international airport in Chisinau (Frontex, 2012, 2013a, 2013b). CPs along the land border were typically set up for a year, but in all cases, their mandates got extended well beyond that, while the CP at the airport was not limited to a certain timeframe (Frontex, 2019a, 2019b). As such, the latter was in operation continuously from its establishment to early 2022 except for a 4-months suspension in 2020 due to pandemic-related restrictions (Frontex, 2021; Interviewee #10).

Officers of the standing corps wore their own uniforms during deployment (Frontex, 2020a), and stayed for about 1–3 months at land border CPs (Interviewee #6), and on ‘long-term deployments’ at the airport CP (Frontex, 2019a, 2019b). They worked under the command of the host country and were thus integrated de jure into domestic chains of command. Between 2020 and 2022, the land border crossing points between Moldova and Ukraine were managed in cooperation by the two countries. Therefore, standing corps officers deployed there were employed by and thus placed under either Moldovan or Ukrainian command, depending on the operational needs and agreements made between the two countries (Frontex, 2011, 2012, 2019a, 2019b).³³

After the signing of the status agreement between the EU and Moldova in 2022, remaining CPs were transformed into Focal Points,³⁴ enabling for the deployment of the standing corps with executive powers. The first deployment took place in March the same year to assist with border control considering the large refugee flows from Ukraine. By mid-2022, the contingent reached 70 officers,³⁵ with plans for it to reach 84 officials.³⁶ Put under the command of Moldovan authorities, these officers got involved in tasks related to border checks, screening, fingerprinting, the registration of migrants, detection of cross-border crime, detection of document and identity fraud, and the support of people with particular needs and vulnerabilities, as well as those in need of international protection’ (Interviewee #10). As opposed to CPs, all officers deployed to FPs along the border between Moldova and Ukraine were put under the command of and received instructions from Moldovan authorities only. On some occasions, Moldovan authorities authorised standing corps staff to use executive powers even in the absence of local officers. Accordingly, operational cooperation at FPs entailed the carrying and use of weapons and other equipment by the standing corps (Frontex, 2022a; Interviewee #10).

Naturally, all joint activities were based on local ownership. While participation in Frontex-led projects was voluntary, the establishment of each CP and FP was requested by Moldova, and Frontex only provided support when it was asked to do so (Frontex, 2011, 2012, 2019a, 2019b). In addition, the evaluation of activities remained largely with domestic authorities, while Frontex and Schengen states relied on the information thus provided to them (Interviewees #3, #9). Moreover, the lifespan of joint controls along certain crossing points between Moldova and Ukraine lasted beyond the involvement of the EU in that these projects got re-established on a bilateral basis while EU experts gradually withdrew. In 2017, the two countries signed bilateral agreements for joint controls at two crossing points – near ‘Pervomaisc – Cuciurgan’ and ‘Giurgiulești – Reni’ – and struck an agreement on related exchange of analytical and statistical data. At the same time, once the EU’s presence at the Palanca–Maiaki-Udobnoe crossing point got re-established via an FP, its operation was also put on a bilateral basis with Frontex-coordinated standing corps having integrated into Moldovan chains of command (Guvernul Republicii Moldova, 2016, 2017; Inspectoratul General al Politei de Frontiera, 2018).

In addition, having learned from its experience with CPs and FPs, on 6 October 2017, Moldova started negotiations on joint controls with Romania. Supported by a Frontex-led study visit in 2018 in the context of the EaP-IBM project for officers from both countries to the Budomierz border crossing point jointly controlled by Poland and Ukraine, Moldova and Romania agreed on coordination in surveillance along the shared border, including via joint investigations and the exchange of information (Frontex et al. 2018a; Interviewee #3).^{37 38}

Finally, with over 70% of deployed standing corps staff having consisted of officers from Schengen states in the studied period,³⁹ cooperation in this context offered another good example of Frontex acting primarily as a facilitator of cooperation between national authorities (Frontex, 2011, 2012, 2019a, 2019b). Indeed, as a Frontex official highlighted, support for regional cooperation between Eastern Neighbourhood and Schengen states as well as domestic inter-agency collaboration ‘was one of the hidden aims of the [EaP-IBM] project’ (Interviewee #9).

Then again, as the above examples demonstrate, joint activities provided ideal frameworks for cross-fertilisation in the form of cross-loading. As a Frontex report on a joint operation concluded, ‘[these activities] created a sense of collegiality amongst [involved international staff, and] enabled [the drafting of] common reports and comparable operational results,’ as well as joint evaluations of the ‘migration situation’ (Frontex, 2019b). This aspect was further emphasised by a Frontex official who stated that Moldovan officers were often asked ‘to provide [Schengen] host authorities with their own expertise on migratory phenomena and regarding documents issued in their home countries or in a language they spoke.’ The same official highlighted that standing corps officers deployed to Moldova ‘observed and took note of local practices in order to widen their professional cultures and awareness of border management’ (Interviewee #10). Although feedback thus gained by the EU from Moldovan officials did not necessarily change practices applied within the EU, it directly influenced Frontex’ – and by large, the EU’s – approach to other third countries (Frontex, 2022a).

In sum, Frontex-led joint activities of Moldovan, EU and SAC border authorities involved a comprehensive pooling of resources with parties increasingly relying on data and know-how provided by their counterparts while integrating into each other’s chains of command at times in a de facto, while other times in a de jure manner in the context of joint activities. In so doing, local ownership on Moldova’s end remained intact with Frontex having largely acted as a facilitator between national authorities and at times vanished in its mediation. Finally, cross-fertilisation via cross-loading of data and know-how formed a key part of this engagement.

Conclusion

The article provided original empirical evidence on EU sectoral integration with Moldova in border management through concentrating on the engagement between Frontex and domestic border and customs authorities. First, it developed a novel conceptual framework for the study of EU sectoral integration. It then applied this to the study of the Frontex-facilitated engagement between EU, SAC and Moldovan border authorities. As the data demonstrates, engagement between these parties remained below the threshold of differentiated integration in that it did not involve the organisational inclusion of Moldovan authorities into EU structures along full convergence to related EU standards. Yet, it reached beyond functional cooperation – generally associated with the Neighbourhood – through the establishment of a shared epistemic community, and a comprehensive pooling of resources.

In addition, this meso-level engagement – adding up to operational integration – between the parties enabled local ownership even if the macro-level cooperation context foresaw full convergence. At the same time, it allowed for cross-fertilisation which involved the uploading and cross-loading of third-country preferences. The EU – via Frontex – acted as a facilitator of transgovernmental engagement between national authorities rather than a supranational entity, and as a hub for the promotion of EU preferences rather than a central regulator, while partially ‘vanishing in its intervention’.

As such, the presented findings on Frontex-Moldova relations reveal a combination of dynamics which set it apart from known engagement approaches sought by the EU in the Eastern Neighbourhood and beyond. While differentiated integration in its entirety has so far only been demonstrated vis-à-vis Western European states, EU trade relations with Moldova, Ukraine, Georgia are found to have come relatively close to it (Müftüler-Baç & Luetgert, 2016). The existing, legally binding ‘deep integration’ between the parties involves the almost complete absence of import duties for 99% of the trade value along the lines of a 95% strong convergence with EU standards thus leaving no room for cherry picking (Bruszt & Langbein, 2017). On the other hand, a good example for functional cooperation in the Eastern Neighbourhood is engagement in the field of environment marked by the provision of limited EU assistance towards facilitating selective convergence (Buzogány, 2013), while none of the countries in the area are even among the ‘cooperation countries’ of the European Environment Agency.⁴⁰ The article thus contributes to the broadening and refinement of the existing conceptual framework of European integration.

Importantly, however, the above findings on sectoral integration should also be studied through the lenses of critical literature. Externalisation – or ‘Europeanisation’ – seen to have largely been exercised by the EU in the Neighbourhood has raised serious ethical concerns and provoked contestation within and outside the Union. On the one hand, it has rendered the study and practice of EU external relations Eurocentric (Keuleers et al., 2016). On the other, it has triggered resistance in ‘strong’ international actors such as Azerbaijan demanding inclusivity and to be treated as ‘an equal’ (van Gils, 2017), and within EU member states such as Hungary seeking de-Europeanisation in various sectors (Müller & Gazsi, 2023). This has led to calls for a re-thinking of EU external relations through finding ways for its ‘decentring’ (Fisher Onar & Nicolaïdis, 2013).

More particularly, sectoral integration based on local ownership and cross-fertilisation might provide an avenue for what is called by scholars of decentring the ‘reconstruction’ of external praxis on the basis of mutuality and local empowerment, a concept which has so far remained rather ‘elusive’ as to its practical applicability (Wolff et al., 2022). In this context, it is worth critically addressing the limitations of sectoral integration, such as the question of equality regards the pooling of resources when it comes to differences between third countries’ rights at de facto integration into Schengen chains of command as opposed to the de jure integration of Schengen staff into third country authorities while enjoying executive powers. Or whether there are even more efficient ways for cross-fertilisation to ensure stronger mutuality.

Acknowledgements

I am immensely thankful to Adam Fagan, Sarah Birch and Sarah Wolff for their invaluable support throughout my PhD project, and to King’s College London for its generous PhD grant. I would also like to thank the reviewers for their constructive comments towards improving the article, and JEPP editor Jeremy Richardson for his kind help along the review process.

Disclosure statement

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

Correction Statement

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

Additional information

Notes on contributors

David Gazsi

David Gazsi is PhD candidate at King’s College London and works as Editorial Assistant with the journal East European Politics (Taylor & Francis).

Notes

1 https://www.eeas.europa.eu/eeas/about-european-external-action-service_en#8419, Accessed 011.08.2022.

2 The ENP is a foreign relations instrument, the Eastern Partnership initiative of which covers six countries, namely Armenia, Azerbaijan, Belarus, Georgia, Moldova, and Ukraine.

3 These include the Border Police (Poliția de Frontieră) and the Customs Office (Serviciul Vamal).

4 ‘Border management’ – used in line with the terminology appearing in EU policies – is understood to include both ‘border control’ through law enforcement authorities, and the work of customs authorities.

5 The EU member states Bulgaria, Cyprus, Romania, and Ireland were not part of the Schengen Area in the studied period.

6 While Norway and Iceland joined the Schengen Area in 1996, Switzerland became a member in 2004, and Liechtenstein in 2008.

7 https://eeas.europa.eu/archives/docs/association/docs/agreements_en.pdf, Accessed 02.07.2022.

8 https://eeas.europa.eu/archives/docs/association/docs/agreements_en.pdf, Accessed 02.07.2022.

9 https://www.eca.europa.eu/Lists/ECADocuments/SR16_09/SR_MIGRATION_EN.pdf, Accessed 08.07.2022.

10 https://ec.europa.eu/commission/presscorner/detail/en/ip_22_7411, Accessed 02.09.2021.

11 https://www.iom.int/eastern-partnership-panel-migration-mobility-and-integrated-border-management, Accessed 02.07.2022.

12 https://frontex.europa.eu/about-frontex/key-documents/?category=working-arrangements-with-non-eu-countries, Accessed 12.07.2021.

13 https://frontex.europa.eu/media-centre/news/news-release/frontex-renews-cooperation-plan-with-moldova-MCXu65, Accessed 02.03.2022.

14 https://frontex.europa.eu/partners/non-eu-countries/, Accessed 05.04.20202.

15 https://ec.europa.eu/home-affairs/policies/schengen-borders-and-visa/border-crossing/eurosur_de, Accessed 02.07.20202.

16 https://frontex.europa.eu/careers/standing-corps/about/, Accessed 02.07.2022.

17 Next to the representatives of EU member states and Schengen-associated countries, the European Commission also sends delegates to the MB.

18 http://eapmigrationpanel.org/sites/default/files/files/discussion_paper_ibm_en_final_updated.pdf, Accessed 02.07.20202.

19 https://www.border.gov.md/index.php/ro/frontex, Accessed 02.10.20202.

20 Accessed 13.06.2018, http://eubam.org/what-we-do/integrated-border-managment/, Accessed 11.07.2020.

21 Accessed 12.06.2018, http://www.eap-ibm-capacitybuilding.eu/en/about/project, Accessed 02.08.2022.

22 The EaP-IBM project was implemented by Frontex in cooperation with ICMPD, the World Customs Organization (WCO), and the International Organization for Migration (IOM).

23 https://neighbourhood-enlargement.ec.europa.eu/system/files/2017-03/enpi_2013_c2013_8293_rap_east_partii_ibm_capacity_building.pdf, Accessed 03.07.2019.

24 https://frontex.europa.eu/media-centre/news/news-release/frontex-concludes-four-year-eastern-partnership-project-YRfU0z, Accessed 10.09.2022.

25 For instance, approximately 82 Moldovan customs officers of different ranks attended trainings in the context of the EaP-IBM project (Interviewee #9).

26 See http://eapmigrationpanel.org/sites/default/files/files/discussion_paper_ibm_en_final_updated.pdf, Accessed 02.07.20202.

27 https://www.border.gov.md/ro/despre-politia-de-frontiera, Accessed 02.07.2022.

28 https://www.border.gov.md/ro/despre-politia-de-frontiera, Accessed 02.07.2022.

29 https://frontex.europa.eu/we-know/situational-awareness-and-monitoring/monitoring-risk-analysis/, Accessed 03.03.2021.

30 https://frontex.europa.eu/assets/Publications/Risk_Analysis/EaP_RAN/EaP-RAN_Q1_2018.pdf, Accessed 02.07.2022.

31 https://frontex.europa.eu/media-centre/news/news-release/frontex-and-moldova-host-international-conference-on-ukraine-and-border-security-d6qZo6, Accessed 08.07.2020.

32 Operations included i.e., Jupiter, Flexi Force, Minerva, Hermes, Alexis, Poseidon, Land, Triton, Hermes, Flexible Operational Activities, Joint Action Days, etc.

33 https://frontex.europa.eu/careers/standing-corps/about/, Accessed 02.10.2022.

34 This transformation took place in the context of the ‘JO Moldova 2022’ project.

35 https://frontex.europa.eu/media-centre/news/news-release/frontex-strengthens-support-to-moldova-43sSie, Accessed 02.10.2022.

36 https://frontex.europa.eu/media-centre/news/news-release/frontex-sending-standing-corps-officers-to-moldova-8KKC9T, Accessed 02.07.20202.

37 https://www.border.gov.md/ro/sefii-politiilor-de-frontiera-din-republica-moldova-si-romania-pledeaza-pentru-dezvoltarea-cooperarii-bilaterale, Accessed 02.07.2022.

38 https://www.border.gov.md/dezvoltarea-cooperarii-dintre-institutiile-de-frontiera-din-moldova-si-romania, Accessed 02.07.2022.

39 https://frontex.europa.eu/careers/standing-corps/about/, Accessed 02.07.2022.

40 https://www.eea.europa.eu/en/countries?size=n_10_n&filters%5B0%5D%5Bfield%5D=readingTime&filters%5B0%5D%5Btype%5D=any&filters%5B0%5D%5Bvalues%5D%5B0%5D%5Bname%5D=All&filters%5B0%5D%5Bvalues%5D%5B0%5D%5BrangeType%5D=fixed&filters%5B1%5D%5Bfield%5D=issued.date&filters%5B1%5D%5Btype%5D=any&filters%5B1%5D%5Bvalues%5D%5B0%5D=Last%205%20years&filters%5B2%5D%5Bfield%5D=language&filters%5B2%5D%5Btype%5D=any&filters%5B2%5D%5Bvalues%5D%5B0%5D=en, Accessed 03.03.2021.

Appendix: List of interviews

Table

Interviewee #1:	Official of the Mission of the Republic of Moldova to the European Union, interviewed online on September 22nd, 2019.
Interviewee #2:	Official of the Directorate-General for Neighbourhood and Enlargement Negotiations of the European Commission, interviewed online on September 25th, 2019.
Interviewee #3:	Official of the Moldovan Customs Service, interviewed on October 1st, 2019, in Chisinau, Moldova.
Interviewee #4:	Local official of the Moldovan office of ICMPD, interviewed on October 2nd, 2019, in Chisinau, Moldova.
Interviewee #5:	Official of the Delegation of the European Union to the Republic of Moldova, interviewed on October 2nd, 2019, in Chisinau, Moldova.
Interviewee #6:	Official of the International Cooperation Directorate of the Moldovan Border Police on duty at the international airport in Chisinau, interviewed on October 3rd, 2019, in Chisinau, Moldova.
Interviewee #7:	Official of the training unit of the Moldovan Border Police, interviewed on October 3rd, 2019, in Chisinau, Moldova.
Interviewee #8:	Official of the Moldovan Border Police deployed to the international airport in Chisinau, interviewed on October 3rd, 2019, in Chisinau, Moldova.
Interviewee #9:	Official of the training unit of Frontex, interviewed online on November 2nd, 2019.
Interviewee #10:	Official of Frontex, interviewed online on June 13th, 2022.