‘Formal informality’ in EU external migration governance: the case of mobility partnerships

ABSTRACT External migration has become a site for extensive EU activity, not least in the years following the start of the migration ‘crisis’ in 2015–2016. It has also become a testing ground for forms of governance that are not generally associated with external relations. We term this type of governance ‘formal informality’ as there is the appearance of an EU agreement with a third country but lacking the legal protections or transparency for actions taken under it. The informality associated with tools of governance is potentially problematic when used to bypass the substantive and procedural formalities associated with law, and which in turn ensure transparency and the protection of rights. This article uses the emergence and evolution of Mobility Partnerships between the EU and third countries as a prism to analyse what the moves towards informality mean for the governance of external migration.

happening in EU migration governance is often ad hoc, not always visible or transparent and couched in terms of 'pilots' and 'projects'. 'Formal informality' here is defined as the appearance of formality, insofar as resembling familiar or established tools (Regulations, Directives, international agreements), but lacking the procedural safeguards, transparency and classification provided by law and legal processes. The 2016 EU-Turkey Statement (which is not an 'agreement' in the legal sense) is a prominent example. Innovations outside legal frameworks, particularly those which exist to provide a format for enhancing relations with third states, might benefit from the flexibility of informality, but need to prompt a change in the way we observe and analyse tools of governance and their transparency. At the heart of migration governance are individuals who are endowed with rights derived from international and EU law, that the EU is Treaty-bound to uphold. In both these consequences, the 'backstop' of the law, underpinning procedural certainty and transparency, risks being lost.
The language of 'crisis' permeating the agenda on migration since (at least) 2015 has led to frenetic searching for urgent solutions (Niemann and Zaun 2018;Niemann and Zaun 2023;Haferlach and Kurban 2017). Emphasis has been placed on engaging the EU's neighbours in migration governance beyond formal tools (such as readmission agreements) with Turkey and Libya, offshore reception centres, and projects to strengthen border controls and disrupt/prevent migration routes. A succession of 'Pacts' agreed by the Member States, most recently in November 2020, seek to cover 'the different elements needed for a comprehensive European approach to migration' (European Commission 2020). The emergence of new ways of governing migration has increased in pace, but a focus on the contemporary overlooks longer trends in external migration governance (Guiraudon 2018). The post-'crisis' period is important, but not the defining factor in the emergence of formal informality. The triggering of the (hithero unused) Temporary Protection Directive as a result of the invasion of Ukraine in February 2022 demonstrates the use of existing EU law in this specific situation, but does not undermine the moves towards innovation in migration governance more widely.
To explore longer-term changes, we analyse the EU's Mobility Partnerships (MPs) established with third countries since 2007. MPs feature strongly in the Pacts and the EU's Global Approach to Migration and Mobility (GAMM) (2005,2011). They have been supplemented by innovations with certain countries, such as Migration Compacts and the Migration Partnership Framework (Seeberg and Zardo 2020). We argue MPs represent a prominent, but hitherto unfamiliar example of 'formal informality' in migration governance. Though the MPs are visible and the product of negotiation with a third state (hence the 'formal'), their content points to a host of initiatives, projects and cooperation mechanisms that are not underpinned by legal certainty (the 'informal'). Through analysis of MPs' evolution, we show how formal informality has moved to a central place in migration governance, alongside agendas and narratives of exclusion and securitisation (D'Amato and Lucarelli 2019).
Taken together, there are significant consequences for our understanding of how EU law and policy interacts with the subject towards which it is directed, i.e. individuals. Without the backstop of the procedural aspects of law (including the possibility of raising legal challenges), there is danger that migration governance has a detrimental effect on the values the EU is committed to ensuring and promoting, including the rule of law, human rights and the respect for international law. The EU's emergence as a key player in global migration governance makes the latter a key concern.
The article first explains the legal structure and framework of EU migration governance and how this framework is being departed from via 'formal informality'. We consider the rationale and emergence of MPs, before tracing examples during different phases of their evaluation since 2007. The conclusion considers the overall impact of the growth of 'formal informality' and what it tells us about Europe's role in global migration governance.

The legal framework of migration governance
External migration is a relative latecomer to EU law. Since the competences granted in the Treaty of Amsterdam 1998, incremental changes have placed regular and irregular migration more fully within mainstream EU law-making. As per Article 79(1) TFEU, the EU 'shall develop a common immigration policy aimed at ensuring, at all stages, the efficient management of migration flows' and provides a legal basis for readmission agreements with third countries.
Early emphasis was placed on familiar methods of legal integration. The use of Directives for the approximation of national laws relating to the Common EU Asylum and Migration Policy was evident in the Tampere Programme (Council of the European Union 1999). The programme kickstarted the search for coherence between national and EU migration policies as the focus shifted from the competence to the substance of migration law (Peers 1998;Dickson 2020). Regulations and Directives passed 1999-2004 dealt with returns, third-country family members and minimum standards in the reception of asylum seekers. Regular migration legislation, such as the 'Blue Card' Directive (2009) has been slower (Kalantaryan and Martin 2015). For some, the 2015 crisis was a direct indicator of the flaws in the EU's institutional set-up and commitments (Menéndez 2016). Overall, the legislation appears to reinforce 'fortress' Europe, a characterisation which has existed for at least 20 years (Peers 1998), and 'externalisation' of the border and migration control via a process of 'shifting up and out' (Lavenex and Uçarer 2004;Lavenex 2006).
Meeting Treaty-based goals via legislation can be fraught with difficulty. The legislative programme already showed signs of stalling by 2002 (Peers 2002). Even with majority voting in the Council, the risk of gridlock is high. With the high political salience of migration and its closeness to core state sovereignty, Member States are often unwilling to further the Treaty-based goals in the spirit of cooperation and solidarity intended in the text of the Treaties. It might seem strange that the Member States allowed the EU extensive competence in migration, only for subsequent legislation to be less than forthcoming. However, the language of 'crisis' suggests that the gradual, familiar path of EU integration has been overtaken by events. Much of the legislation referred to above was completed before the crisis. The establishment of a Common European Asylum System (CEAS) marked the dawn of the 'age of implementation' (Tsourdi 2020). Costello's (2020) view of the CEAS is, however, that it 'is destined to remain a policy failure, an ethical sore, and political tinderbox'. Not all Member States, whether new or old, have proved ready to commit themselves (publicly) to enforceable legislation. Hungary and Slovakia unsuccessfully challenged compulsory refugee quotas at the Court of Justice. Member States risk not respecting the provisions of EU law in migration, especially against the background of a highly politicised area where 'Brussels' is often the scapegoat (Goldner Lang 2020). The same can be said for the process of concluding international agreements where political, legal and practical reasons prevent their conclusion with a third country (Ott 2020).
Furthermore, the European Parliament has frequently flexed its muscles to avoid the danger that its powers are not circumvented in migration competence, including in litigation before the Court of Justice (Sánchez-Barrueco 2020, 63). The Parliament has traditionally prided itself on upholding rights and vocal about the use of legislation for repressive means, although a long-term trend towards voting with the Council for a more restrictive position on irregular migration has been noted (Acosta 2009;Lopatin 2013). For international agreements with third countries, particularly where there is only a limited pre-existing relationship, there is also likely to be a high level of scrutiny on the EU side. For an agreement to take effect, the third state must also agree, and there may be a host of reasons why they might not (see Chou and Gibert 2012). Both these factors can make the informal approach, bypassing higher levels of scrutiny, attractive for Member States and third countries alike.
The lack of appetite for legislation appears to have impacted on the forging of a strategic approach, which has gradually downplayed the core role of legislation. The GAMM (2005GAMM ( , 2011 emphasises the undefined use of 'all available tools' and a 'full range of migration instruments' (Cardwell 2013(Cardwell , 2018. These all aim to further the Treaty goals in some way, but do not rely on the law as we understand it, implying that the procedural safeguards and oversight of the Court do not apply. Yet, it is important to not be overly optimistic about the capacity of law to guarantee conformity with procedural standards, transparency and human rights. Legal obligations do not in themselves prevent unfairness or unlawfulness. A readmission agreement with a third country may not prevent (or even could be said to facilitate) returns where there is a risk of harm to an individual by removing the procedural safeguards. However, operating outside established frameworks, especially in 'crisis-mode', can lead to unpredictable outcomes and 'new style of discretionary governance' (Joerges and Kreuder-Sonnen 2016;Scicluna 2018). The drift towards informality has allowed Member States to 'create alliances outside the constraint of the EU legal and institutional framework' and 'unhook themselves from formulating a common EU migration and asylum policy' (Panizzon and van Riemsdijk 2019). This drift is particularly pertinent given the lacuna in international law and responsibility for actions that EU institutions and agencies (particularly Frontex) operate in (Cortinovis 2015;Sánchez-Tabernero and Santos Vara 2016;Carrera and Cortinovis 2019;Giuffre 2020;Loschi and Slominski 2021). Legal processes help to understand where, e.g. 'humanitarian' measures are more security-related within an uncertain legal/governance space (Moreno-Lax 2018; Cusumano 2019; Bello 2020) despite the clarity of international-derived obligations including non-refoulement (Mungianu 2016) and dangers of circumvention (Farahat and Markard 2020). This moves the inquiry beyond the preliminary question of whether the EU or Member States are responsible for external action (Garcia Andrade 2018), to what they produce and potential conflicts with Treaty-based values (Bauböck 2018;Murray and Longo 2018) including the EU's self-identification as a 'beacon' for international human rights (Barbulescu 2017).
The move away from traditional legal instruments can be explained by the newer forms filling gaps, but allowing opportunities for flexibility and shielding from democratic processes associated with law's formality. Herein lies theusefulness of the terminology of 'governance': its descriptive qualities identify a large(r) set of institutions involved in the interface between law-making, policy-making and administrative practices. In turn, the use of governance (as opposed to simply 'law') necessarily indicates a shift in thinking about the kind of things that matter in decision-making, legal-institutional dynamics and departures from 'traditional' paradigms. The language of governance is commonplace, but often comes with a normative association with 'good' governance. The inquiry here, which examines whether the emergence of instruments that deal with migration and the potential detriment of (legal) rights and due process, is anything but 'good'. The use of governance in the migration context therefore uncouples it from use in areas where new forms of democratic participation and inclusion can occur.
The use of 'formal informality' here is designed to capture a particular kind of governance, that is to say, where the appearance of formality is provided by the similarity of instruments to established forms (such as international agreements) but which appear to fall short in accounting for exactly what actions are taken underneath them, and what is their legal or rational connection to other instruments. MPs stand out as malleable examples that have been used over a substantive time period. MPs might be considered as an example of non-binding soft law (Ott 2020; Tittel-Mosser 2020). However, we do not use this term here since a category of soft law does not adequately capture the degree to which MPs have developed a dual character as instruments in their own right, and facilitators of relationships and processes having practical effects. Similarly, referring to MPs as non-binding and lacking legal effects negates the potential for the creation of process-based legal effects that should take place under the formality of (legal) agreements. We therefore use 'formal informality' to situate MPs somewhere between 'hard' and 'soft' but with particular characteristics that have developed over time.

Mobility partnerships in EU migration policy
MPs emerged against the background of the changes in governance explored above, and the shifting of the EU's post-enlargement external borders. The European Neighbourhood Policy (ENP), concluded in 2004, provided a framework for deepened relations with new and old neighbours, overlapping with existing frameworks with Mediterranean and Eastern neighbours (Cardwell 2011). MPs further the migration cooperation aspects of the ENP, which gradually came to the fore in the bilateral dimensions with partner states (Eisele and Wiesbrock 2011;Weinar 2017).
MPs exist with most ENP states, and Cape Verde. EU migration strategies, including the GAMM (2011), European Agenda on Migration (European Commission 2015) and New Pact on Migration and Asylum (2020) feature MPs as a governance 'tool'. GAMM emphasises the start of mobility 'dialogues' with individual ENP states in the aftermath of the Arab Spring (Carrera, Parkin, and den Hertog 2013;Dandashly 2015). As per the GAMM, 'The dialogues allow the EU and the partner countries to discuss in a comprehensive manner all aspects of their possible cooperation in managing migration flows and circulation of persons with a view to establishing Mobility Partnerships'.
MPs represent 'formal informality' for three reasons. First, they are more than peripheral statements of intent. They occupy an increasingly central place in bilateral relations with selected third countries, and over a significant period (since 2007). The Commission describes them as 'the most complete framework for bilateral cooperation between the EU and its partners, based on mutual offers of commitments and project initiatives covering mobility, migration and asylum issues'.
Second, they are relatively 'concrete' insofar as there is an agreed text. They provide a vaguely-defined gateway for practical cooperation but require commitment from both sides. Tittel-Mosser (2023) explores the factors for third countries in deciding whether to agree to an EU Mobility Partnership, which has also contributed to the strong differentiation in the bilateral relationships (Okyay et al. 2020). MPs appear therefore as 'lawlike': stopping short of concrete or enforceable measures but nevertheless capable of producing material effects. As an 'umbrella' for cooperation, they allow dynamic patterns of cooperation between some Member States only with the third country via practical cooperation (Frasca 2021). Though beyond the scope of this article, the use of informal mechanisms between individual Member States and third countries is significant (see, for example, Reviglio 2020 on the Memorandum of Understanding between Italy and Libya).
Third, they typify the kind of emerging governance in this area: they do not require pre-existing legal relationships and are subject to high levels of differentiation depending the partner country concerned. In the absence of a set template or list of protections or safeguards, their vagueness can become a site of intense discursive struggles (Kunz and Maisenbacher 2013) as different policy priorities are thrashed out between the EU Member States as well as with the third state.

Mobility partnerships between migration and foreign policy
The stated purpose of MPs is to deepen migration cooperation with a third state on a 'coowned' basis. All MPs cover regular and irregular migration: to 'facilitate' migration (often 'circular and temporary migration') alongside 'genuine cooperation on migration and development', but also 'preventing and combating illegal immigration, smuggling of migrants and trafficking in human beings'. The incentive for partner states is linked to the possibility of opening routes for 'regular' migration, including visa facilitation (Maroukis and Triandafyllidou 2013). Assistance with upgrading both national laws on migration and furnishing the partner state with equipment to reinforce migration control is usually foreseen in the MPs.
MPs tend to cover similar ground, suggesting that they are built around EU preferences and not the tailored needs or interests of the diverse partner states (Lavenex and Stucky 2011;Moreno-Lax 2020). They are open-ended in terms of cooperation. Support for the development of the partner state (Lavenex and Kunz 2008) and its capacities are, from the text of the agreements at least, crucial. GAMM (2011) itself refers to 'numerous projects supporting the implementation of Mobility Partnerships'. Tittel-Mosser (2023) identifies the existence of bilateral ties with Member States, the strength of negotiating power of the third country and its level of administrative capacity as guiding the content of MPs. Similarly, Stutz (2023) finds multiple motivations on the part of the EU for cooperation, which do not primarily rely on the number of migrants from, or transiting through, the partner state. Zardo (2017) argues that countries in transition have less bargaining power due to their need for international legitimacy and domestic survival, particularly for post-Arab Spring governments (e.g. Tunisia). Others argue that third countries are not passive recipients of policy and involvement is the result of cost-benefit calculations (Reslow 2012;Reslow and Vink 2015). This has provoked new dynamics in migration diplomacy, whereby third countries have gained significant concessions from the EU (Laube 2021, Vaagland 2023. Some, including Senegal and Egypt, have decided that the political costs of a MP are unacceptable and negotiations have not progressed (Chou and Gibert 2012). This alone suggests MPs have obligations and costs that cannot be dismissed as meaningless.
MPs also serve as an interface with other aspects of foreign policy, such as the aim to 'mainstream migration issues into key development sectors' (European Commission 2015; Zaun and Nantermoz 2023) and raising complex inter-institutional questions with security and defence (Bergmann and Müller 2023) and development (Cortinovis 2017). Over-simplified categorisation suggests that migration law is 'real' law, based on enforceable legal texts or international agreements, whereas foreign policy is 'political' and outside the scope of legal analysis. A MP purports to be both: facilitating measures which have effects akin to legal effects, whilst being an initiative that is not traceable to a specific legal competence. The overlapping of foreign policy priorities with migration signals the possibility of a legal grey area that does not rely on traditional Regulations and Directives but a place where innovative and non-binding/non-enforceable measures appear. The increased securitisation of migration policy does not necessarily indicate the EU is failing to deliver on its policy objectives, rather it could be interpreted as expansionist governance which seeks out new ways to enhance migration policy at the EU level, thus furthering integration (Boswell, Geddes, and Brumat 2020).
Scholarly opinion on MPs has veered between a cautious welcome for their cooperative nature and joint ownership, and criticism for being 'driven by a strong conditionality requiring third states to adopt and show strong commitment to European security policy on mobility as irregular migration' (Carrera, Parkin, and den Hertog 2013) and their lack of clarity and coherence (Parkes 2009). It is acknowledged that MPs strike an unequal bargain in terms of what can be gained for a third state (Maisenbacher 2015) but the Commission is also in a precarious situation when it comes to offering incentives, as these largely rely on Member State commitments. Hence, the lack of transparency about what the EU or the Member States can offer with MPs has tended to place them within the context of the securitisation of migration, which has received much criticism (Seeberg 2017;Bello 2020).
Discussing success in terms of goal realisation can be premature as the MP can be a launchpad for multiple initiatives, or simply remain dormant (Reslow 2015). Their use might be prioritised in times of crisis and dependant on identifying routes for preventing irregular migration, with evidence of differentiation between the various partners, even if the texts share common provisions. But that is not the focus here. Rather, it is that MPs as an example of formal informality in migration governance provide a gateway to practical cooperation shielded from wider scrutiny. The evolution of the MPs in force suggests that a transition has occured to a framework foregrounding cooperation on preventing irregular migration whilst offering less of a route to a formal agreement for regular migration routes and visa facilitation.

The evolution of mobility partnerships and 'formal informality'
Twelve MPs have been concluded since 2007. Their content vary, which reflects the different criteria identified by Tittel-Mosser (2023) that have guided their emergence. A chronological tracing shows how their relationship with formal, legal instruments, including readmission agreements, has evolved. Since the MPs are a product of bilateral negotiations, even the GAMM (2011) decried their 'uneven relevance and importance' and lack of 'clear logical interrelationships'. Here, we divide MPs into phases to demonstrate their changing relationship with wider EU migration strategies and the consolidation of formal informality in migration governance. Selected examples are set out for each of the phases.

The 'pilot phase'
The GAMM (2011)  For Moldova, the stated purpose of the MP is to 'facilitate' mobility and 'well managed migration' for the mutual benefit of Moldova and the EU. This discourse is also apparent in the GAMM, where MPs are presented as an opportunity for mobility rather than a preventative measure. Thus, the MP confirms the commitment of both to improving the movement of persons whilst also 'working to ensure better management of migration flows, including preventing and reducing illegal migration flows'.
A cooperation framework is supported by 'more information, integration and protection for migrants' as well as a reduction to the risks of illegal migration and trafficking of human beings, but little indication of how this is operationalised. The process by which the MP is agreed differs from the Association Agreement in that there are no formal negotiations. Instead, a visa dialogue commenced in 2010 and within that context an Action Plan in 2011. The latter specified the indicators for reform that would determine Moldova's progress towards EU norms and the MP. Four areas were identified: document security, irregular migration, public safety, and foreign relations and fundamental rights. However, as Guérin and Rittberger (2020) have noted, Moldova was already embarking on domestic policy-change and modernisation in migration law and policy. The motivation of pursuing aims under the MP is unclear. Buracec's analysis (2012) of the impact of the MP on EU-Moldova relations reveals that it opened further possibilities for bilateral cooperation, and institutional contacts via the key stakeholders identified within the agreement. Twinning projects in border management envisage an enhanced role for Frontex.
The EU-Georgia MP was signed in 2009 with the VFA and RA entering into force in 2011. The focus is similar to the Moldovan MP with strong emphasis on facilitating mobility. Preventing and reducing irregular migration flows appears a secondary concern: The Mobility Partnership will have the purpose of better managing legal migration including circular and temporary migration … preventing and combatting illegal immigration and trafficking in and smuggling of human beings, as well as promoting an effective readmission and return policy, while respecting human rights and the relevant instruments for the protection of refugees. (European Commission 2009) Further, the MP is 'conceived as a long-term framework based on political dialogue and cooperation' and 'will evolve over time'. This emphasises the flexible nature of MPs to respond to developments and reward progress on the one hand, but also opens the possibility for a rebalancing of measures focussing on irregular migration.
The language of this MP is less affirmative in terms of providing a concrete and defined framework, it speaks to 'promoting' a better framework, which indicates that cooperation remains open-ended. There is an emphasis on institutional capacity-building, including provisions on asylum policy so Georgia can implement its own policy and manage caseby-case regularisation under national law. As with Moldova, there are provisions 'to inform potential migrants on ways of legal migration to the EU and on requirements for legal staying' (Council of the European Union 2008). This, of course, does not require creating migration opportunities but only to publicise any that exist.
More details are provided on implementation, specifically 'the capacity to monitor migration through the improvement of data collection and analysis systems and cooperation with countries of destination'. Although specific information on implementation lacks transparency, Moldova and Georgia are considered by the European Commission as the 'best pupils' in the implementation of their MPs. Brouillette (2018) considers this led to a 'common understanding' between the EU and national actors that may lead to a translation of the European objectives in the field of migration into the registries of practices in the neighbourhood.

The 'maturing' phase
Following the pilot phase, the extension of MPs to other ENP states include those with Armenia and Azerbaijan. The EU-Armenia MP was signed in 2011 (European Commission 2011) with a VFA and RA in 2012 and 2013. The EU-Armenia MP maintains a continued emphasis on the benefits of migration. In promoting a better framework for legal and labour mobility, the first provision names temporary and circular migration as key initiatives. In its first article, the Armenian MP eschews the vague, open-ended nature of the earlier MPs. Instead, a preference for 'concrete and effective initiatives' is articulated: To promote a better framework for legal and labour mobility, including through the facilitation of temporary and circular migration, supported by more information and concrete and effective initiatives as well as protection of migrants. (European Commission 2011) A striking feature of the EU-Armenian MP is the prevalence of information in informing the development of the relationship. To this end, Article 2 establishes the enhancement of migration monitoring as a key dimension. Types of monitoring relationship are envisaged, including information to be shared among relevant agencies. The implementation of the MP is subject to a more concrete institutional framework for bilateral cooperation, perhaps a response to earlier criticisms that MPs had not been fully implemented (Reslow 2015), but nevertheless remaining unspecific in terms of respective responsibilities or how the partner state will be involved. For instance, the European Training Foundation is central in the delivery and co-ordination of pre-departure training, and Frontex for the development of coordinated border policy. Despite the focus on information sharing and improved coordination across the levels of government, Article 22 stresses the need to avoid overlap in activities and ensuring the 'cooperation platform' is used to incorporate relevant actors. This demonstrates an awareness that the MP promotes institutional contacts but leaving the limits to cooperation are and at what point informal contact becomes institutionalised cooperation unclear.
The development of cooperation is based on risk analysis, explicitly to the 'fight against irregular migration and related cross border crime' (Article 11). Combatting irregular migration relies on cooperation between EU border guard organisations of the EU and Frontex 'as and when appropriate'. This raises several concerns with regards jurisdiction and legal responsibility, particularly in relation to international legal protections for refugees and asylum seekers.
Azerbaijan signed an MP and VFA in 2013 (European Commission 2013). The RA followed in 2014. The partnership outlines four purposes, and as with Armenia, the focus on managing labour migration includes reference to temporary and circular migration as well as labour market and socio-economic conditions. There is a more explicit reference to international standards and the protection of human rights, particularly in relation to preventing and combating irregular migration and maximising the development impact of mobility. Unlike the EU-Georgia MP, this does not include the purpose of establishing an effective asylum system but does state as a purpose the promotion of international protection and the enhancement of its external dimension. This is an interesting development that might operate as a precursor to the MPs concluded more recently that emphasise the role of external third countries in the 'shared' management of asylum claims and refugees. The capacity to monitor migration is once again a central focus of the partnership, including the development and maintenance of country migration profiles, inter-agency cooperation and information exchange. Article 19 outlines that dialogue and consultation between the partners will determine the issues of 'common interest' and needs linked to migration.
The EU-Azerbaijan MP is 'a long-term co-operation framework in line with the GAMM and based on political dialogue and cooperation' and to be developed on the basis of existing agreements such as the Partnership and Coordination Agreement (European Commission 2013). The MP does state that it 'will evolve over time' which emphasises its open-ended nature and leaves the choice of specific governance mechanisms open to change depending on circumstances. This characteristic has become more pronounced in more recently concluded MPs.

'Next generation' mobility partnerships
More recent MPs have emerged against the backdrop of the migration 'crisis' as a focal point of the EU's agenda (Scipioni 2018;Baldwin-Edwards, Blitz, and Crawley 2019). We argue that there has been a shift in the content and development of MPs as a result of the crisis and its impact. The 'next generation' phase builds on existing MPs but mobilises them to new ends. Examples are found in North Africa, where bilateral EU partnerships are generally weaker, but which have been the focus of action on irregular migration. Such examples show how cooperation has been reinvigorated to respond to migration challenges the EU faces and move MPs beyond their original scope.
The most developed MP example is Tunisia. Domestic democratisation in Tunisia in 2011 resulted in the offer of an EU 'privileged partnership' (2012) and an ENP Action Plan that enhanced political contacts and financial assistance for broad, cross-cutting reform agendas. The MP was agreed in March 2014 and subsequent negotiations for a VFA and RA commenced in 2016.
As a 'next generation' MP, the EU-Tunisia MP has a different focus. After signing the Joint Declaration establishing the MP, EU Commissioner for Home Affairs, Cecilia Malmström, stated it 'aims to facilitate the movement of people between the EU and Tunisia and to promote a common and responsible management of existing migratory flows'. The suggestion of bi-directional movement masks the shift in the use of MP as a tool of crisis management. Emphasis is placed more squarely on the EU's expectations of the partner in tackling problems and sharing responsibility.
The 'next generation' MPs have however been less forthcoming in terms of mobility to the EU, and more focused on mechanisms for returns and readmission of migrants. The tying of the MP to a formal RA appears to have vanished. Hence, the formal aspect that the MP has acted as a bridge towards is no longer present, with informality at the core of the relationship. VFA negotiations since 2017 to facilitate Tunisian mobility to the EU have been slow. The EU-Tunisia Association Council in 2018 agreed strategic priorities for the period 2018-2020. VFA negotiations are not explicitly mentioned, although the creation of 'future prospects for young people' is identified as a priority alongside 'technically skilled' migration. Tunisia has not implemented, in the EU's view, necessary reforms to make progress as the VFA is being negotiating in parallel with a RA. For this, the EU wants Tunisia to accept returns of Tunisians but also third-country nationals who travelled via Tunisia. Tunisia's reluctance to agree to this stipulation has prevented the VFA. Even tying the VFA as an important step towards a deeper free trade agreement has not resulted in swift progress.
Others with relatively strong EU bilateral relations are moving in a similar direction. Morocco became an EU 'privileged partner' in 2004, and following its MP (Council of the EU 2013), saw an increase in funding via the Emergency Trust Fund for Africa (see Zaun and Nantermoz 2023) and European Neighbourhood Instrument 2014-2017. However, negotiations have not yet commenced on a VFA or RA, so cooperation appears less formalised than with Tunisia (see Cassarino 2014).
The four objectives are similar to other MPs. Instead of temporary migration, the term 'short periods' is used but the emphasis remains on (changeable) labour market dynamics. The focus of the development objective is the exploitation of the 'potential of migration and its positive effects' but remains vague. In contrast, for irregular immigration the objective is clear: to combat it and promote effective returns and readmission. A final objective is compliance with international instruments concerning refugee protection. What follows is a broad programme of how these objectives will be met.
The MP with Morocco offers few concrete outcomes. Despite the active phrasing, the MP remains highly open-ended and flexible in nature. Provision 32 commits the parties 'To consolidate cooperation and dialogue between the parties in order to adopt a common approach to migration' but does not detail the process for consolidation. Provisions pertaining to the rights of individuals are also open-ended, e.g.: to improve coherence between the policy concerning the mobility of persons and other areas of sectoral cooperation (trade, education, research, culture) for Moroccan vocational trainees, students, academics, researchers and businessmen and women to enter and stay in EU Member States. (Provision 7) (Council of the EU 2013) The choice of 'endeavour' signifies any gains are only potential and progress may not transpire. Morocco has long resisted the signing of a readmission agreement (El Qadim 2018;Wolff 2014) which would result in a potential large number of returns, not only of Moroccan citizens but those who have transited en route. And yet, despite this gap, the EU-Morocco MP has facilitated relations to the extent that is holds one of the largest portfolios of cooperation with the EU. The Moroccan example is thus instructive in the decoupling of a formal RA as the logical consequence of concluding an MP but the development of informal cooperation, particularly on irregular migration.

Furthering 'formal informality' via the migration partnership framework
Relevant to the discussion is the more recent Migration Partnership Framework (MPF), introduced in June 2016. The MPF serves to fully integrate migration into EU foreign policy. The purpose of the MPF is to mobilise the instruments, resources and influence of both the EU and its MS to establish cooperation with partner countries to 'sustainably manage migration flows'. Another priority is better cooperation with a view to identifying, re-documenting and readmitting nationals from these countries (Slominski and Trauner 2018). Initially the agreements under the MPF were envisaged as 'compacts', akin to the EU-Turkey Statement and the EU-Jordan Compact. Such compacts would establish clear commitments from each partner around which other EU policy areas could be mobilised to deliver on migration goals. The MPF is now described as a 'political framework for continued and operational cooperation, pulling together the different work strands in order to develop a comprehensive partnership with third countries, combining instruments, tools and leverages available to the EU and MS to deliver clear targets and joint commitments' (European Commission 2016). This appears to remove any immediate benefit for partner countries, or at least ones that are publicised, since there might be benefits hidden via this demonstration of 'formality informality', particularly which suit mutual (and hidden) securitisation agendas (Seeberg 2017). The leverages referred to point to the pressure on third countries to re-admit their nationals and those of other countries who have used the third country as a transit point. Leverages might also include financial incentives, which might help to explain the need for secrecy. What remains, however, are benefits for the EU and the Member States in this approach: Well-designed, flexible and streamlined instruments to address migration challenges will be necessary in the coming years. In the past, rigid rules on programming of resources have been an obstacle to responding rapidly to MS' requests, or lead to a systematic recourse to (limited) emergency funding. There are also areas where EU funding instruments could play a stronger role in the future. (Commission 2017) Therefore, despite lack of incentive on the part of third-countries, the EU seems committed to pursuing this style of governance for the future. In addition, the EU has re-purposed existing MPs with strategic countries to further assist its migration management policies, with Morocco and Tunisia as prominent examples of strengthening practical cooperation but with less emphasis on formal legal measures, given the difficulties in creating them. As Reslow (2015) has found, the MPF significantly departs from MPs. Much less attention is devoted to RA and VFA, which had been front-and-centre of the earliest examples. In fact, the MPF makes no mention of formal agreements on visa facilitation or readmission as a priority. Rather, the aim on facilitating 'fast and operational returns' and 'specific and measurable increases in the number and rate of return and readmission' suggests that the use of the frameworks is revealing in terms of the emphasis placed on practical cooperation, even in the absence of formal agreements. The role of individual Member States in engaging in bilateral cooperation, with the EU as semi-detached coordinator or facilitator appears to characterise the MPFs. The reference to funding mechanisms is also telling and offers an insight into how the participation of partner states can be secured, even as the EU and its Member States have shown themselves to be less willing to enhance regular migration opportunities. The existence of the MPF supports the analysis here that the drift towards informality continues in migration governance.

Conclusions
Mobility Partnerships concluded between the EU and its neighbours over a ten-year period, pre-and post-'crisis', offer both an insight into the evolution of migration governance and an opportunity to see how the lines between formality and informality have become blurred. By characterising MPs as an example of 'formal informality', this analysis demonstrates that migration governance operates on the basis of initiatives that appear to have the legitimacy, both procedural and substantive, of formal agreements, but which in fact do not undergo the same processes. Their flexibility and vagueness mean that there are few opportunities to discover the practical levels of cooperation that are facilitated by such tools, which in an area where there are vulnerable people and their rights at stake, is troublesome (Moreno-Lax 2020). The strong dimension of differentiation between the partners and undefined means of cooperation has become a hallmark of the haze under which migration governance has developed. This is where the importance of terminology is found. At first glance, the MPs are 'soft' instruments: statements of intent and a means of signalling engagement on an important issue of governance, both European and global. They fulfil the EU's Treaty-based obligations to forge close(r) relationships with neighbourhood countries and, on the face of it, constitute a give-and-take approach to migration which balances the interests of both. Yet, the evolution of the MPs demonstrates that it is the interests of the EU and its Member States at the forefront of their content. Whilst this might not be surprising, given the EU's larger bargaining power, the growing emphasis on combatting irregular migration and the avenues the MPs open in doing so represent something new, and concerning. The power imbalance supports a turn to measures which resemble law but without the crucial characteristics associated with law. As Costello (2020) has put it, 'When the EU sought to control the refugee arrivals in 2016, it turned to extra-legal means, which so far have withstood legal challenge'. Whilst attention has been focussed on the EU-Turkey Statement, this applies equally to the MPs but with the added difficulty of pinning down any specific provisions which might fall foul of EU or international law.
The shifts in language that have followed the MPs through the different stages of their evolution are thus significant. Presenting MPs as a political framework for a bilateral relationship, building on a pre-existing (general) agreement and backed up by (formal) agreements on readmission was the original intention. However, this has been replaced by a refashioning of MPs towards crisis management. Therefore, the formal dimension appears to have been dropped. This matters, since the legal underpinning is no longer there as the baseline for the EU's governance of migration. Further still, locating the instances of cooperation under the 'umbrella' of the MP is challenging, especially since interactions may be at the level of one of more Member States with the partner state. Thus, even tracing the budgetary spending cannot reveal the extent of this informality.
What this tells us about the EU's migration governance of migration, and by extension its contribution to global migration governance, is that there is a continuing preference for engaging third countries in preventing irregular migration before migrants can arrive in Europe, and providing the financial and physical means to do so. The MPs furthermore serve as a template for formal informality with countries located further from the territory of the EU, but which the EU has identified as potential partners in migration. Hence, the vagueness of the language associated with MPs can also be found in Migration Compacts (Seeberg and Zardo 2020) and the Common Agendas on Migration with Nigeria, India and Ethiopia. All these follow the same path of cooperation, but without the pre-existence of formal, legal agreements . Indeed, the recognition of CAMMs as pre-cursors to MPs is identified in the GAMM itself in such terms, where 'one side or the other is not ready to enter into the full set of obligations and commitments'. In this respect, due regard should be given to the commitments the EU has given itself in the Treaties, namely 'the protection of human rights, [and] the strict observance and the development of international law, including respect for the principles of the United Nations Charter'. Judged on these criteria, the implications of a shift to informality, especially with tools that resemble the formality of law, have serious consequences and considerations for the EU, its neighbours, global migration governance and our approaches to understanding.