Impact of the Agent’s Environment on Discretion in the Field of EU Conflict Resolution

  • Friedrich Plank
  • Arne Niemann
Open Access
Part of the Palgrave Studies in European Union Politics book series (PSEUP)


Employing an agent–principal perspective on EU conflict resolution and focusing on the politics of discretion, we argue that the context in which the agent acts affects the principal–agent relation. Our conceptualization draws on the agent’s actions as a source of discretion, which can be either an unintentional product of the characteristics of the external environment in which the agent performs (structure-induced discretion) or a product of intentionally pursued agent action (interest-induced discretion). Focusing particularly on the European External Action Service, we find that the agent can exploit urgency and third party interplay as characteristics of EU conflict resolution in order to increase his discretion. This has led to an agent that has increasingly sought to further its interests and accumulate powers that affect discretion.

1 Introduction

Confronted with failed states, ongoing civil wars and violent conflict in the neighbourhood, the European Union (EU) increasingly engages in conflict resolution. While a broad range of actors representing the EU in managing conflict have been examined, little research has been conducted on actor relations in EU conflict resolution. In this chapter, we broaden the use of the principal–agent model in the study of EU external action and shed some light on the relations and political dynamics that underlie EU conflict resolution. 1 We particularly argue that the context in which the agent acts affects the principal–agent relation in this field. Whereas principal–agent research on the EU is no longer in its infancy, a number of important issues merit further analysis.

Overall, it seems that while substantial efforts have been made to investigate the principal side of principal–agent relations (Brandsma and Blom-Hansen 2016; Conceição-Heldt 2011; Delreux 2008; Epstein and O’Halloran 1999; Leblond 2011), additional research is necessary on the agent side of the relationship. EU conflict resolution is a policy area characterized by non-specific “acts of delegation”, an external context demanding urgency and a complex policy environment with various third parties. For these reasons, it constitutes a relevant field for analysing “agent–principal” relations in the EU.

Despite the fact that principals (here the member states) want to be in control of EU conflict resolution due to specific interests or colonial ties, the complex character of violent conflicts often requires a high degree of flexibility and urgent action of the agent (here the European External Action Service, EEAS). In addition, EU conflict resolution is implemented within a complex system of organizational overplay involving various third parties such as the United Nations (UN), regional organizations as well as third countries (Motsamai and Brosig 2014). The impact of this complex environment on agent discretion is interesting and relevant: Does the described characterization of EU conflict resolution provide structural conditions for a high discretion of the agents and are they thus “European agents out of control” (Klein 2010, title) or does it rather limit the agent’s discretion? Examining the agent side in EU conflict resolution, we focus on the following question: How and under which conditions are agents able to acquire more discretion vis-à-vis their principals and to what extent does the external environment of EU conflict resolution contribute to this acquirement of additional discretion?

The principal–agent relationships and the theoretical focus are visualized in Fig. 1. We conceptualize the agent’s actions as a source of his discretion, which can be either an unintentional product of the characteristics of the external environment in which the agent performs its task (structure-induced discretion) or a product of intentionally pursued agent action (interest-induced discretion) (Delreux and Adriaensen this volume). While actions taken by the principals (such as control) affect agent discretion, agent-related sources may constitute additional important factors for discretion vis-à-vis the principals. Drawing on the principal–agent literature and conflict resolution research, we operationalize conditions for both structure-induced and interest-induced agent discretion.
Fig. 1

Visualization of the principal–agent relationship and the impact of the agent’s environment on discretion

We argue that in EU conflict resolution, agent discretion is largely structure-induced. It constitutes a policy domain which is characterized by high degrees of urgency and interplay with third parties. These structural conditions are particularly felt on the agent side of the dyad and facilitate discretion-enhancing behaviour. However, in order to get a complete picture of agent discretion in EU conflict resolution, we also need to take interest-induced discretion into account, as agents tend to maximize their own interests and thus alter those of the principals. They can do so by exploiting their agenda-setting powers, such as chairing of meetings and drafting of proposals, as well as leveraging the information advantage they enjoy vis-à-vis the principals.

We aim specifically to fill two gaps in the existing literature: (1) focusing specifically on the agent in principal–agent relations, our approach primarily contributes to improving our knowledge of the conditions of agent discretion in a complex policy environment shaped by various actors; and (2) by investigating EU conflict resolution through the lens of the principal–agent model, we link two important strands of literature. While there are now quite a number of analyses of EU foreign policy from a principal–agent perspective (e.g. JEPP special issue 2011), conflict resolution has barely been touched upon in previous studies. Since we also assess conflict resolution outside the European Neighbourhood Policy (ENP), we broaden existing research which mostly focuses on conflict resolution via the EU as a framework or via EU policies that refer for instance to conditionality (Hill 2001; Smith 1998; Tocci 2007). 2

2 Theoretical Puzzle: The Agent’s Actions as a Source of Discretion

The delegation of tasks to agents implies not only benefits but also potential costs. To avoid these costs, principals create control mechanisms which “define […] the scope of agency activity, the legal instruments available to the agency, and the procedures to be followed by it” (Pollack 2003, 27). While principal–agent scholars have distinguished between ex ante, ex post and ad locum control mechanisms (Delreux 2010; Elsig 2010), our approach focuses on the agent’s discretion vis-à-vis the principal, which may be either structure-induced or interest-induced (Delreux and Adriaensen this volume). Previous findings indicate that the agent’s discretion is influenced not only by internal but also by external explanatory variables (Delreux 2013). Studies have found that the context of negotiations (Delreux 2010), the strength of the framework of an international organization (Billiet 2009) and the informality of the institution the agent is involved in (Niemann and Huigens 2011) increase the agent’s discretion.

In general, the degree of discretion depends on the actions of the principals with regard to the amount of authority that is initially delegated as well as the control mechanisms imposed by the principals. However, discretion is not the inverse of control. We assume that agents are also able to enhance their discretion vis-à-vis the principals—which constitutes the main difference between autonomy and discretion. The latter also includes the autonomy captured by the agent (Delreux 2013).

Agent discretion may thus arise as a product of the structure in which the agent performs its task and/or as a result of agent actions, that is to say, structure-induced or interest-induced discretion. The former refers to the fact that discretion not only is affected by the control mechanisms imposed by the principals, but can also be the result of the external environment in which the agent fulfils his tasks. This concept is related to “agency slippage” which “occurs when constraints or incentives provided by the principals induce the agent to behave in ways systematically different from those preferred by the principals” (Pollack 2003, 26). However, in contrast to “slippage”, structure-induced discretion refers to specific pressures which are a product of the environment in which the agent operates. Thus, our understanding of structure-induced discretion is broader than the traditional idea of “slippage”, since it also includes the structure in which the agent performs its task as source of the agent’s discretion. Most importantly, these pressures are only felt on the agent side of the dyad; principals which frequently are not directly involved in action “on the ground” are excluded from the external structures affecting the agent’s fulfilment of the delegated tasks. The external environment can shape the behaviour of the agent and thus also affect the agent’s discretion (Delreux and Adriaensen this volume).

We specifically assume two characteristics of the external environment that can affect the agent’s discretion. First, the specific context of the policy domain might mandate a high degree of urgency and complexity of EU actions, which in turn requires significant flexibility for the agents. Since dynamics in conflicts and crises are often swiftly changing, flexible and urgent action of the agent is imperative (Delreux and Keukeleire 2016). Second, third party behaviour is another determinant of actor discretion falling under the structure-induced umbrella. Increasingly, EU foreign policy takes place in cooperation with other actors. These include interregional fora, international organizations and third countries. In this context, the agent might also have an informational advantage since the organizational interplay requires coordination (Elsig and Dupont 2012). As principal–agent scholars have already found (Delreux 2009; Delreux and Kerremans 2010), negotiations with third parties can produce certain pressures for the agents to move beyond their mandate.

Interest-induced discretion, conversely, is related to “shirking” which according to Pollack (2003, 26) “occurs, when an agent pursues preferences of its own rather than, or to the detriment of, the preferences of the principals”. This concept assumes the agent to be self-interested and pursuing his own agenda (Hawkins et al. 2006; Hawkins and Jacoby 2006; Kiewiet and McCubbins 1991). In contrast, discretion which is interest-induced also includes actions taken by the agent to alter the preferences of the principals. These actions can weaken the principals’ propensity to activate control mechanisms, e.g. the likelihood to police patrol meetings, to select specific agents or to limit the budget. Agents can increase their discretion by exacerbating the problem of hidden information for the principals (Hawkins and Jacoby 2006; Kiewiet and McCubbins 1991). This information asymmetry often involves institutional memory on the part of the agent. Moreover, strategies might have an effect on the discretion the agents enjoy, for instance when they interpret their mandates broadly in the context of cooperation and negotiation with third parties (Delreux 2009) or by building support and setting specific focal points during negotiations with other third parties (Elsig and Dupont 2012). As our analysis will show, the scope for interest-induced discretion is often related to characteristics of the structure in which agents act. Agents might make use of the external environment and the structural setting around the principal–agent relationship in order to maximize own preferences and to alter those of the principals. We identify the agent’s powers in drafting proposals and chairing meetings as important agenda-setting strategies that can be intentionally exploited in order to increase the discretion of the agent. Moreover, information asymmetry between agent(s) and principals facilitates the pursuit of self-interests. How far agents can pursue their self-interest strongly depends on the characteristics of the environment in which EU conflict resolution is implemented.

3 Empirical Analysis: The EEAS’s Discretion in EU Conflict Resolution

Our empirical section is structured as follows. As a first step, we map the concerned principal–agent relations in EU conflict resolution as an important precursor to our analysis. We then point to characteristics of EU conflict resolution as a policy domain and illustrate how the external environment enables the agent to increase its discretion vis-à-vis the principals. The third and final part of this section illustrates the drivers for growing agent discretion in EU conflict resolution.

3.1 Imprecise Mandates in the EU Conflict Resolution Principal–Agent Relationship

In the field of EU conflict resolution, the principals, defined as actors that both grant and have the power to revoke authority (Hawkins et al. 2006), are the member states of the EU. After the entry into force of the Lisbon Treaty, the main agent is the EEAS (see Fig. 1). Important actors within the EEAS include the High Representative (HR), the EU Special Representatives (EUSRs) as well as the EU Delegations, the latter with different roles before and after Lisbon Treaty. We treat the EEAS as a monolithic agent and hence consider the EU Delegations, the EUSRs and the High Representative as representing an undivided EEAS. Our conceptualization of the EEAS as a monolithic agent is based on the assumption that principals delegate specific tasks—such as the representation of the EU in third countries—to the EEAS that coordinates these activities by including sub-actors into a web of delegation. We assume that the fulfilment of tasks is then in the hands of the EEAS as a whole. For analytical purposes, and in order to decrease the complexity of the act of internal delegation, delegation within the EEAS is not discussed in our analysis. The fulfilment of the task by the actors included in the EEAS is governed by the same act of delegation (see also Gastinger this volume; Laloux this volume). With regard to the EUSRs, we assume that internal delegation within the EEAS has considerable effects on their mandates since they are proposed by the High Representative in the Council.

While the domain of EU conflict resolution includes a wide range of tasks delegated, we focus in particular on two core tasks: (1) the implementation and coordination of engagement abroad, e.g. missions and operations; and (2) the diplomatic representation of the EU in third countries and multilateral organizations. The two tasks that are part of the act of delegation seem to be rather relevant in this regard.

First, concerning EU crisis management through military operations and civil missions, the legal framework of the EU emphasizes the leading role of the EEAS for EU crisis management after the Lisbon Treaty (Dijkstra this volume). Crisis management is conducted and coordinated within the EEAS Crisis Response and Operational Coordination Department, which was created by High Representative Ashton at the end of 2010. It is supported by a Situation Room and the EEAS Crisis Platform which serve as coordination bodies for all relevant EU engagement in a crisis including information sharing. The EEAS is clearly placed at the forefront of a comprehensive approach to EU conflict resolution (Tercovich 2014).

Second, with regard to the diplomatic representation of the EU in third countries, in particular the EU Delegations and the EUSRs as part of the EEAS are relevant to consider. While the Lisbon Treaty states that “Union [D]elegations in third countries and at international organisations shall represent the Union” (Article 221 TFEU), the role of the EU Delegations, in particular with regard to the representation of the EU at international organizations, faces numerous problems of its staff acquisition and overlapping competences (Helwig et al. 2013; Dijkstra this volume; Helwig this volume). The tasks of the Delegations can be summarized as follows: they ensure the external representation of EU foreign policy with third countries and multilateral organizations. In addition, they organize and host regular meetings (at least monthly) at the level of Heads of Mission. They also lead coordination at the technical levels. In general, EU Delegations conduct political dialogue and ensure diplomatic representation on behalf of the EU (Helly et al. 2014). However, the specific powers of the EU Delegations depend on the quantity of the staff and the working relationship with relevant member states (Helly et al. 2014; Interview K) as well as the quality of the Delegations networks and contacts in the host country (Interviews G, I, J).

The staff’s network and their expertise also matter for the EUSRs who serve as implementing actors of EU conflict resolution since 1996. On the basis of a proposal by the High Representative, they are appointed by the Foreign Affairs Council which also extends and terminates their mandates. The EUSRs, who have been depicted as the EU’s “ears and eyes” (Grevi 2007: 11), provide information about the mandated area, contribute to the formulation of EU policy towards their mandated area and to the coordination with third parties and stakeholders, represent the EU in specific forums such as the Middle East Quartet and maintain contacts and dialogue in the region (Tolksdorf 2014). However, some mandates are more detailed than others (Interview H). For instance, the mandate extension of Alexander Rondos, EUSR for the Horn of Africa, specifies coordination with regional organizations, contribution to peace and stability in Somalia, Sudan and South Sudan as well as representation in multilateral fora and engagement in conflict resolution as tasks (European Union 2015a). Some EUSRs are located in Brussels, whereas others execute their task in the mandated area. While it is said that Baroness Ashton was rather sceptical towards the EUSRs because of their (perceived) member state affiliation (Interview G), Frederica Mogherini has adopted a more open attitude towards the EUSRs (Fouéré 2016).

The High Representative is the most visible actor within the EEAS. Her role has been enhanced substantially through the entry into force of the Lisbon Treaty by “double-hatting” the post and making its occupant both a member of the Commission and the chair of the Foreign Affairs Council (Edwards 2013). The Lisbon Treaty specifies her tasks as follows: “The High Representative shall represent the Union for matters relating to the common foreign and security policy. He shall conduct political dialogue with third parties on the Union’s behalf and shall express the Union’s position in international organizations and at international conferences” (Article 18 TFEU). Encouraged to contribute to crisis management with own proposals, the High Representative is considered as a rather autonomous player with enhanced agenda-setting powers (Vanhoonacker and Pomorska 2013; Helwig this volume).

3.2 The External Environment of EU Conflict Resolution: Urgency and Third Party Interplay

The previous section has shown that the field of EU conflict resolution is characterized by rather imprecise mandates in terms of specific actions to be carried out by the agent. The precision of mandates in a more general sense is often dependent on the conflict’s own characteristics. Two major characteristics define the environment in which EU conflict resolution takes place: (1) urgency and flexibility of action and (2) the interplay with third parties.

EU conflict resolution commonly constitutes a reaction to the escalation of violent conflicts. It only provides short windows of opportunity to contribute to conflict management processes. In many cases, these specific circumstances are urgent and require a high degree of flexibility of the agents. The EU’s involvement in Aceh/Indonesia is a case in point (Interviews F, I). After a Tsunami hit the region in 2004, the risk increased that tensions between the conflict parties—the Indonesian government and the Free Aceh Movement or Gerakan Aceh Merdeka (GAM)—would escalate. Both parties agreed to negotiate as they understood there was high time pressure. When an agreement was finally signed in 2005, it became apparent that the vacuum of one month between the signature and the establishment of the EU-led Aceh Monitoring Mission (AMM) could jeopardize peace. Hence, the Council Secretariat and its Crisis Management Directorate again acted rapidly as it drew up a concept for an Initial Monitoring Presence (IMP) to cover the month. “A clear mandate by the Council was missing” (Braud and Grevi 2005: 22). These urgent actions of the agents which included both the Commission and the Council Secretariat 3 were crucial for the start of the AMM since the staff of the IMP could provide for expertise once the AMM started. Flexibility of the agents was also a factor during the negotiations of the MoU, which were mediated by the Crisis Management Initiative led by former Finnish President Martti Ahtisaari. In fact, Ahtisaari maintained time pressure as he set straight deadlines for the conflict parties during the mediation (Interview I).

In general, the environment in which the EEAS implements EU conflict resolution is frequently characterized by urgency and time pressure. Thus, quick action as well as a high degree of flexibility, for instance with regard to adapting to dynamic situations such as a coup, is often emphasized (Pirozzi 2013; Tocci 2007). In mediation activities and civil as well as military missions, the EEAS is exposed to time pressure. The Belgrade–Pristina dialogue which started on 2011 under the mediation of Robert Cooper, High Representative Ashton’s adviser on the Western Balkans, serves as a good example. When the International Court of Justice ruled that Kosovo’s unilateral declaration of independence was applicable to international law, the High Representative and her office as implementing agent had to adapt to the new circumstances rather quickly (Bergmann 2016; Bergmann and Niemann 2015). As Serbia was circulating an own draft for a UN resolution calling for renewed talks on the status of Kosovo, there was enormous time pressure (Economides and Ker-Lindsay 2015). Another example can be derived from High Representative Ashton’s engagement during the crisis in Egypt after President Morsi had been removed from office. Ashton managed to quickly launch a diplomatic mission and was the first to have direct access to Mohammed Morsi in his detention facility (Morillas 2015; Pinfari 2013). The High Representative could build on previous efforts in 2011 and 2012. Both the EU Delegation in Cairo and the EUSR for the Mediterranean, Bernardino León, were perceived as swift independent peace brokers and “lesser evil” than others (Morillas 2015: 26).

The interplay with third parties constitutes the second major characteristic of the environment in which the EEAS fulfils its tasks. Most often, EU conflict resolution takes place in a multi-actor environment that is shaped by the interplay between various organizations such as the UN, regional organizations and third countries. The agent fulfils his tasks in dense institutional spaces where multiple actors are involved pursuing their own agendas. In Aceh, for instance, the AMM which monitored the implementation of the peace agreement involved countries from the Association of Southeast Asian Nations (ASEAN) and served as a preliminary test for interregional security cooperation. The EU’s efforts in the Central African Republic (CAR) after inter-ethnic violence broke out in late 2013 were part of a complex web of institutional overlay between the EU and various regional initiatives (Welz 2016). When EUFOR RCA was deployed in spring 2014, coordination with the African-led International Support Mission to the Central African Republic (MISCA) was of particular importance for an effective reduction in the proliferating violence. The EU mission concentrated on specific areas around the capital Bangui and thus facilitated the presence of MISCA in remote parts of the country. The mandate which focused on stabilization, provision of humanitarian assistance as well as support of mediation efforts by the African Union (AU) was clearly dependent on coordination with third parties (Plank 2017).

Institutional interplays are also important from a more general perspective which emphasizes the overall importance of interaction with international partners. Often, regional efforts serve as a basis for EU engagement (Interview B), and many interviewees reiterated the importance of regional empowerment and ownership as a general aim of EU foreign policy (Interviews B, D, K). The EU’s engagement in the Mali peace talks in 2013 illustrates the importance of a supportive role for regional efforts. While the EUSR for the Sahel, Michel Dominique Reveyrand-de Menthon, contributed substantially to the achievement of the Ouagadougou Accord of June 2013, his role as a supporter of regional efforts by the ECOWAS and the AU was clearly expressed in the mandate (Davis 2015). In Yemen, conflict resolution by the EEAS through the EU Delegation was embedded in a group of ten different actors, which included countries from the Gulf Cooperation Council and the permanent members of the UN Security Council (Girke 2015).

3.3 Drivers for Growing Agent Discretion in EU Conflict Resolution

The external environment in which the EEAS fulfils its task generates possibilities for the agent to increase its discretion vis-à-vis the principals. Since member states often pursue own interests in the management of conflicts, such as France’s increased engagement in former colonies, the principals want to retain firm control in EU conflict resolution which affects their security interests significantly. In order to increase their discretion, agents resort to discretion-enhancing strategies that might make use of the external environment.

3.3.1 The Impact of Urgency and Flexibility of Action on Discretion

The EEAS as main agent implementing EU conflict resolution is equipped with flexible tools to react to urgent crises in particular through its crisis platform (Tercovich 2014). For instance, the EAAS can act immediately through the EUSRs when there is a crisis or dynamics which require urgent action (Interview A). While not all EUSRs are permanently located in their mandated area, they usually have staff on site and can coordinate and act quickly through their networks. The EUSRs can rely on their skills and experience with quick coordination processes with stakeholders and conflict parties (Interviews G, J). These specific competences might facilitate to present agent action as a fait accompli. The Political and Security Committee (PSC) is then only briefed by the EUSR. While some EUSRs “clean” their reports before discussion (Edwards 2013), urgency can facilitate strategies to increase the discretion of the agent. The capacity to act flexibly and swiftly particularly applies to EUSRs that have a well-developed network and a mandate in a turbulent region. For instance, it has been noted that Alexander Rondos, EUSR for the Horn of Africa, enjoys much more discretion than other EUSRs because of his competence and network in the region (Interview I). Accordingly, the principals agreed to his broad mandate which leaves the EUSR as the main body of EU engagement in the Horn of Africa. The turbulent character of the context in which Rondo is fulfilling his tasks is specifically mentioned in his mandate extension which refers to the potential deteriorating context of the external environment (European Union 2015b).

Particularly in cases of high urgency, the agent has opportunities to set the agenda. After the emergence of the South Sudan crisis in 2013, which took many member states by surprise (Interviews C, H), Rondos, who had intense contacts with the relevant conflict parties and staff located in Addis Ababa, Khartoum and Juba, had several dialogues with the parties involved in the renewed fighting between factions of the government in South Sudan (Interview C). EU conflict resolution was then dependent on the quick action of the EUSRs office, which was available during the Christmas holidays. The principals left the matter to the agent due to the latter’s flexibility to react to the dire situation. Intense contacts to conflict parties had also been used by Pekka Havisto, Rondo’s predecessor, who tried to coordinate the EU’s efforts in a proactive and independent way during the negotiations leading to the Comprehensive Peace Agreement (CPA) in Sudan (Ferhatović 2010; Interview H). These examples illustrate that the EEAS can exploit a high degree of urgency through the EUSRs since they have a considerable lead in implementing quick action.

The EU Delegations are of particular interest in this regard, too. On the ground, the EEAS can make use of the EU Delegations. As one interviewee suggested, Delegations can independently develop action rather quickly, in particular in third countries that are not at the centre of attention of the member states (Interview H). In many (small or less prominent) third countries, the EU Delegations only interact with few national embassies. The principals then leave the agent’s actions uncontested since many principals do not have their own embassies in the third countries and rely on the EEAS as an information provider and first point of contact (Interviews A, E). As the case of the Belgrade–Pristina dialogue illustrates, urgency can serve as “favourable ground” for an increased discretion of the EEAS (Amadio Viceré 2016: 566; see also Vanhoonacker and Pomorska 2013). In other cases, such as the crises in Egypt, the agent can exploit the urgency of the context in order to set the agenda and to expedite conflict resolution efforts which then are implemented by the agent itself such as by Baroness Ashton in the case of diplomatic initiatives after the detention of former President Mursi (Morillas 2015; Pinfari 2013). Ashton’s engagement in the mediation of both conflicts was condoned by the principals. Her (previously built) contacts and quick action facilitated a rather uncontrolled engagement.

3.3.2 The Impact of Interplay with Third Parties on Discretion

Inter-organizational interplay of EU conflict resolution constitutes a major feature of the policy domain. In general, there is a different environment abroad which facilitates agents to coordinate their efforts with third parties (Interview H). As a consequence, the EEAS—in particular through the EUSRs and the EU Delegations to the AU or to the ECOWAS—can leverage this to increase its discretion in discussions with interregional partners (Interviews C, F). The EU’s engagement in Somalia, the CAR, Sudan and Mali has been linked to coordination pressures by the third parties involved. In particular, inter-organizational relations to the UN, the AU and other (sub-)regional organizations require close coordination. These partners—which also include officials from the conflict countries—seek to have an institutional day-to-day partner they can work with (Interviews B, C). The EU Delegations and the EUSRs usually have the mandate and the local presence to do so (Interview F). While the AU–EU relationship is developed and intensified through PSC-to-PSC meetings and joint ministerial summits and thus controlled by the principals to some extent, staff from the EEAS coordinates on a day-to-day basis with the AU, which often seems to be more relevant (Interviews B, C). Interregional relations are specifically implemented by the EUSRs in many cases (Interview A), particularly since meetings such as the PSC-to-PSC rounds focus on broader issues of the interregional partnership, which are then explicitly implemented by the EEAS. One interesting example can be derived from the EU’s cooperation with the Intergovernmental Authority on Development (IGAD). While there is no partnership formalized, the intense coordination of conflict resolution efforts in Somalia, Sudan and elsewhere is implemented by the EUSR for the Horn of Africa, Alexander Rondos (Interview J). Thus, the EEAS has substantial discretion in EU–IGAD relations since there is no formal agreement and control by the principals is rather limited.

The growing cooperation with third parties in the field of EU conflict resolution has substantially increased the discretion of the EEAS in this specific policy area. Day-to-day coordination with partner organizations facilitates growing agent discretion since the partners put pressure on the agent to move beyond the mandate or to interpret the mandate broadly with regard to specific requirements put forward by the third parties. The coordination is the responsibility of the agent and escapes principal scrutiny. The EEAS can then set specific focal points (Elsig and Dupont 2012) such as in the relations with IGAD when for instance the EUSR adapts to the positions of regional organizations towards specific conflicts. The joint engagement of the EU with counterparts in Africa is clearly under coordination pressure, both operationally and administratively. In the case of AMISOM, the AU-led peace keeping force in Somalia, the EEAS is the main point of reference for the AU through the EU Delegation in Addis Ababa. Similarly, the EU Delegation to Somalia enjoys substantial discretion since it serves as “the head financial officer for the aid provided by the EU and acts as National Authorizing Officer for the processing of project approvals” (Rein 2015: 199). The Regional Indicative Programs under the framework of the African Peace Facility that allocate a substantial amount of money to the partners are managed by the EU Delegations. The EEAS has been able to exploit its exposed status as main interlocutor with third parties (Interview A). Control by the principals is then exerted ex ante, and the implementation is left rather uncontested to the agent, which might also be linked to a more practical and cost-effective logic of the principals (Interview H).

In cases such as Sudan, the presence of envoys and delegations of the USA, China, Russia or regional organizations increases the discretion of the EEAS, since they act within a peer group that structurally strengthens the agent abroad (Interview H). The EUSR can then refer to the positions of other important players and link potential strategies and propositions to a comprehensive approach involving regional and international stakeholders. For instance, with regard to international pressure during the conflicts between the Sudan and the South Sudan after the independence referendum in 2011, this comprehensive approach had been emphasized by the EUSR (Interview H). EU Delegations have been able to exploit joint missions with third parties in Mali, Pakistan as well as with the ECOWAS in order to increase their competence. This includes sending a military liaison officer to the Delegation in Abuja (ECOWAS) as well as embedding support functions for operations in the Delegation in Nairobi (Pirozzi 2013). In these cases, the principals were apparently convinced that the agent needed additional discretion and a broader mandate in order to respond to third parties’ requests.

These examples illustrate that the EU Delegations under the umbrella of the EEAS act within a specific web made up of various third parties. In particular, EU Delegations to regional organizations, which increasingly serve as main partners of EU conflict resolution, are able to use their status as main reference points from the perspective of the third parties (Interviews B, C). The complex external environment characterized by urgent dynamics and a web of institutional interplay thus tends to constitute a supportive setting for increased agent discretion.

3.3.3 The Internal Dimension of Increased Agent Discretion

While we identified structure-induced factors as an important source of increased discretion of the EEAS, interest-induced factors additionally affect the agent–principal relations in EU conflict resolution. From an internal perspective on EU conflict resolution, we identify in particular one factor that allows the EEAS to engage in strategies in order to increase its discretion vis-à-vis the principals. The EEAS has substantial power through process-related institutional competences, as staff from the EEAS chair the meetings of the PSC as well as the working groups (Helwig et al. 2013). These are the most important decision-making bodies in EU conflict resolution outside of the Foreign Affairs Council, which is chaired by the HR regarding all issues relevant to EU conflict resolution. 4 This chairing of meetings and the preparation of the agenda is seen as a potential source of increased agent discretion (Interview H). As one interviewed EU official put it: “If you are the first one to put something on paper, then you have already set the tone and the direction of the discussion. And it does happen that member states do not agree and push it in a different direction, but the fact that you have set out the first markers has a huge influence on the discussion” (Interview J). In some cases, staff from the EEAS went to the PSC several times without significantly changing its proposal, for instance with regard to the suspension of sanctions against Belarus, and pushed their agenda through the Committee (Interview E).

The Foreign Affairs Council is also strongly influenced by the agenda-setting powers of the HR. The structure of chairing and preparing grants discretion to the HR, for example with regard to overloading the agenda or to tabling proposals that require member states alliances or the opposition of large member states in order to be blocked (Interview A). If the HR and her team in the EEAS propose, for instance, to extend all the mandates of the EUSRs—as happened reportedly in a Council meeting and several PSC meetings in early October 2015 (Interviews A, J)—this creates an environment in which member state diplomats cannot easily veto the proposal. While the drafting of specific policy proposals is increasingly requested by member states (Interview J), isolating a single member state within the Council meetings is one strategy available for the HR in order to increase discretion. In particular, small member states try to avoid isolation or blocking a decision (Interview G). Chairing and drafting are also relevant with regard to the presence of the EU in third countries since the EU Delegations chair the meetings of the ambassadors and prepare the agenda (Interviews H, I, K). This creates an environment in which the EU Delegations have substantial influence on a comprehensive approach of the EU towards a third country. Overall, we recognize an increased role for the EEAS abroad, for instance with regard to the EUSRs (Interview B). As in the case of Myanmar, agent action in EU conflict resolution can also serve as means to be further perceived as a valuable partner in the region (Interview F).

At the same time, the information asymmetry between agents and principals has substantially grown in recent years, particularly with regard to small member states (Interview B). This confirms previous research. Dijkstra and Vanhoonacker (2011) conclude in their study on EU foreign policy from an informational perspective that the Lisbon Treaty and the creation of the EEAS have further expanded the capacities for autonomous information gathering. In particular, small member states heavily rely on the skills and expertise of the EEAS, which has divisions and staff that work daily on specific areas and issues (Interview B).

Both agenda-setting powers and information asymmetry facilitate an intentional increase in agent discretion. It seems that the EEAS increasingly pursues its own agenda to alter principals’ preferences, which refers to the above-mentioned concept of interest-induced discretion. While the HR has interests in files such as the Ukraine crisis and the SOPHIA mission in the Mediterranean Sea, the EEAS pushed the suspension of sanctions against Belarus after political prisoners had been released by the regime through the PSC in order to conclude the negotiations (Interview E). EU Delegations, which are described as independent bodies with independent agendas (Marangoni and Vanhoonacker 2015; Interview I), have their own interests and networks. Interviewed diplomats conclude that agents want to maximize their own power and to establish themselves as important actors (Interviews D, E, J). The EEAS has increasingly developed an institutional memory (Interview B). In sum, “[a]t the moment we have a strong HR, a strong Commission and an EEAS which is more elaborating on positions” (Interview B).

Two internal factors facilitate increased agent discretion in EU conflict resolution: the chairing and drafting powers of the EEAS as well as an information asymmetry between principals and agent. This interest-induced dimension of agent discretion in the field of EU conflict resolution allows the EEAS to increasingly pursue own interest and to alter those of their principals.

4 Theoretical Argument

The external environment clearly affects the relations between principals and agents of EU conflict resolution. Despite more general ambitions to enhance the newly created institutions, we particularly identify the external environment as a source of increased discretion for the EEAS. The urgency of the context and the interplay with third parties constitute important factors for increased agent discretion. In addition, some institutional changes of the Lisbon Treaty equipped the agent with powers that can potentially be used to increase the discretion of the EEAS vis-à-vis the principals. This internal dimension of strategies to influence the politics of delegation might also give rise to an agent which increasingly pursues own interests and thus acquires interest-induced discretion.

From a theoretical perspective, our core argument emphasizes that the agent’s discretion can be affected by the characteristics of the external environment. Specific strategies that exploit pressures of urgency and third party interplay, which are particularly felt on the agent side of the relationship between principals and agent, then refer to structure-induced discretion. Increasingly, EU foreign policy takes place in a complex environment shaped by urgency and the involvement of third parties. It is embedded in an intense web of coordination and cooperation with virtually every country and most regions in the world. While interaction with third parties is a specific goal of the EU—expressed with the creation of the EEAS—it also creates pressures to adapt to the positions of other actors that are particularly felt on the agent side. Many decisions and actions of the agent additionally require substantial flexibility since the environment in which the tasks are fulfilled is characterized by swiftly changing dynamics. Frequently, foreign policy primarily reacts to the urgency of specific events.

In addition, interest-induced factors that increase the discretion of the agent, such as agenda-setting powers, apparently affect the principal–agent relationship. In particular, the powers of the EEAS in chairing and drafting constitute important “instruments” in this regard. Agents not only pursue their own interests and develop an institutional memory, but can make attempts to alter the preferences of their principals, for instance by creating (perceived) constraints to veto or change proposals put forward by the agent.

In summary, we contribute to the literature on the principal–agent model by specifically focusing on the agent and the external environment as a source for increased discretion. We have also shown that the principal–agent model is applicable to complex policy domains such as conflict resolution and that the complexity of the context might increase agent discretion. In addition, we emphasize that discretion is also affected by agent actions and internal factors that may impact on the range of independent action available to the agent for the fulfilment of their tasks.

5 Conclusion

Starting from the observation that a number of factors affect agent discretion, we have shown that the effects of the external environment of EU conflict resolution (urgency and third party involvement) on the discretion of the EEAS as agent are most noteworthy. In addition, the information asymmetry between principals and agent as well as specific powers of the EEAS (such as agenda setting) facilitate the agent to pursue strategies in order to increase discretion. This has led to an agent that has increasingly developed and sought to further its interests and accumulate powers that affect discretion. In sum, both structure-induced and interest-induced factors that shape agent–principal relations facilitate an increase in agent discretion vis-à-vis the principals in EU conflict resolution. The findings of this study contribute to previous research (Elsig and Dupont 2012; Delreux and Kerremans 2010), which has already mentioned time pressure, agenda setting powers of the agents and the strategies of EU institutions as important factors increasing discretion. Conflict resolution as a research object for principal–agent scholars is thus to some extent similar to the EU’s negotiations with third parties (more generally), in which informality of the institutional environment and agent expertise (Niemann and Huigens 2011), the compellingness of the context, and a large number of third actors (Delreux 2010) have been identified as factors that affect agent discretion. In the complex arena of foreign policy, control by the principals is limited, since they delegate to institutions that have their own agendas and networks with other actors in the international system. Hence, though principals usually have the last say on the EU’s engagement in conflicts, control might potentially entail high political costs for the principals with regard to domestic scepticism on the one hand and increasing security threats, migration flows and wavering international credibility on the other. Our core argument is: agents have the capacity to increase their discretion vis-à-vis the principals in a complex setting, which contributes to opportunistic and self-interested agents in EU conflict resolution. EU external policy then appears as a policy area in which “the notion of ‘delegated, and therefore controlled authority’ […] no longer holds” (Zürn 2012: 85; see also Haftel and Thompson 2006). Our analysis also indicates that some areas of EU conflict resolution are actually characterized by firm control exercised by the principals. While interplay with third parties might facilitate increased agent discretion in EU conflict resolution, there also exist areas in which member states limit delegation to agents. For instance, the discretion of the EEAS with regard to diplomatic representation in Washington and Moscow remains rather limited (Interview D). Also a case such as the Ukrainian crisis has shown that an environmental context of urgency and interplay with various actors does not necessarily create increased agent discretion when vital interests of the member states are at stake.

While EU conflict resolution still remains an under-researched field of EU foreign policy, we suggest that there is substantial ground for future research, particularly by applying the principal–agent model. With regard to the competences of the High Representative (e.g. the fact that EUSRs have to report to the Council via the HR), an assessment of additional principals in EU foreign policy could be another avenue for future research. Moreover, both in-depth case analyses and comparative assessments of relations in EU conflict resolution would contribute to an interesting policy field that increasingly displays relevance.


  1. 1.

    To define the concept “conflict resolution” “[d]ifferent authors and practitioners use […] terms in inconsistent ways” (Miall 2004: 2–3). The typologies include “conflict resolution”, “conflict transformation” and “conflict management”. In contrast to what is often assumed, we define conflict resolution as a separate concept among others rather than an umbrella term. In contrast to “conflict transformation”, which is long-term oriented (e.g. Lederach 1995), it refers to third party policies which aim to settle violent conflict in the short term. The European Commission defines “conflict resolution” as “actions undertaken over the short term to end violent conflict” (Duke and Ojanen 2006: 484). These actions include third party mediation and military interventions as well as EU missions or the prevention of conflicts.

  2. 2.

    Our analysis is based on interviews with EU officials conducted in Brussels and via telephone in autumn 2015. In addition, we assessed official documents and case-specific literature.

  3. 3.

    In the pre-Lisbon period, the Council Secretariat and the Commission were the main agents implementing EU conflict resolution.

  4. 4.

    The importance of the PSC is also perceived by the diplomats of the member states themselves (see Howorth 2010).


Cited Interviews


Member state official

October 2015



Member state official

October 2015



EU official

October 2015



EU official

October 2015



Member state official

October 2015



EU official

October 2015



Member state official

October 2015



EU official

October 2015



EU expert

October 2015



Member state official

November 2015



EU official

November 2015



  1. Amadio, Viceré M. (2016). The roles of the president of the European council and the high representative in leading EU foreign policy on Kosovo. Journal of European Integration, 38(5), 557–570.CrossRefGoogle Scholar
  2. Bergmann, J. (2016). Same table, different menus? A comparison of UN and EU mediation practice in Kosovo. Paper presented at the UACES 46th Annual Conference, London.Google Scholar
  3. Bergmann, J., & Niemann, A. (2015). Mediating international conflicts: The European Union as an effective peacemaker? Journal of Common Market Studies, 53(5), 957–975.CrossRefGoogle Scholar
  4. Billiet, S. (2009). Principal–agent analysis and the study of the EU: What about the EC’s external relations? Comparative European Politics, 7(4), 435–454.CrossRefGoogle Scholar
  5. Brandsma, G., & Blom-Hansen, J. (2016). Controlling delegated powers in the post-Lisbon European Union. Journal of European Public Policy, 23(4), 531–549.CrossRefGoogle Scholar
  6. Braud, P., & Grevi, G. (2005). The EU mission in aceh: Implementing the peace. Paris: European Union institute for security studies.Google Scholar
  7. Conceição-Heldt, E. (2011). Variation in EU member states’ preferences and the commission’s discretion in the Doha round. Journal of European Public Policy, 18(3), 403–419.CrossRefGoogle Scholar
  8. Davis, L. (2015). Reform or Business as usual? EU security provision in complex contexts: Mali. Global Society, 29(5), 260–279.CrossRefGoogle Scholar
  9. Delreux, T. (2008). The EU as a negotiator in multilateral chemicals negotiations: Multiple principals different agents. Journal of European Public Policy, 15(7), 1069–1086.CrossRefGoogle Scholar
  10. Delreux, T. (2009). The EU negotiates multilateral environmental agreements: Explaining the agent’s discretion. Journal of European Public Policy, 16(5), 719–737.CrossRefGoogle Scholar
  11. Delreux, T. (2010). The EU as international environmental negotiator. Burlington: Ashgate.Google Scholar
  12. Delreux, T. (2013). Examining EU external action through the lens of principal–agent theory. Paper presented at the 8th Pan-European Conference on International Relations, Warsaw.Google Scholar
  13. Delreux, T., & Adriaensen, J. (2017). Introduction. Use and limitations of the principal–agent model in studying the European Union. In T. Delreux & J. Adriaensen (Eds.), The principal–agent model and the European Union (pp. 1–34). London: Palgrave MacMillan.Google Scholar
  14. Delreux, T., & Kerremans, B. (2010). How agents weaken their principals incentives to control: The case of EU negotiators and EU member states in multilateral negotiations. Journal of European Integration, 32(4), 357–374.CrossRefGoogle Scholar
  15. Delreux, T., & Keukeleire, S. (2016). Informal division of labour in EU foreign policy-making. Journal of European Public Policy. doi:  10.1080/13501763.2016.1216151.
  16. Dijkstra, H. (2012). The influence of EU officials in European security and defence. European Security, 21(3), 311–327.CrossRefGoogle Scholar
  17. Dijkstra, H. (2017). Non-exclusive delegation to the European external action service. In T. Delreux & J. Adriaensen (Eds.), The principal–agent model and the European Union (pp. 55–81). London: Palgrave MacMillan.Google Scholar
  18. Dijkstra, H., & Vanhoonacker, S. (2011). The changing politics of information in European foreign policy. Journal of European Integration, 33(5), 541–558.CrossRefGoogle Scholar
  19. Duke, S., & Ojanen, H. (2006). Bridging internal and external security: Lessons from the European security and defence policy. Journal of European Integration, 28(5), 477–494.CrossRefGoogle Scholar
  20. Economides, S., & Ker-Lindsay, J. (2015). “Pre-Accession Europeanization”: The case of Serbia and Kosovo: Pre-accession Europeanization. Journal of Common Market Studies, 53(5), 1027–1044.CrossRefGoogle Scholar
  21. Edwards, G. (2013). The EU’s foreign policy and the search for effect. International Relations, 27(3), 276–291.CrossRefGoogle Scholar
  22. Elsig, M. (2010). European Union trade policy after enlargement: Larger crowds shifting priorities and informal decision-making. Journal of European Public Policy, 17(6), 781–798.CrossRefGoogle Scholar
  23. Elsig, M., & Dupont, C. (2012). European Union meets South Korea: Bureaucratic interests exporter discrimination and the negotiations of trade agreements. Journal of Common Market Studies, 50(3), 492–507.CrossRefGoogle Scholar
  24. Epstein, D., & O’Halloran, S. (1999). Delegating powers: A transaction cost politics approach to policy making under separate powers. Cambridge: Cambridge University Press.CrossRefGoogle Scholar
  25. European Union. (2015a). Council Decision (CFSP). 2015/440.Google Scholar
  26. European Union. (2015b). Council Decision (CFSP). 2015(2006).Google Scholar
  27. Ferhatović, E. (2010). Implementing peace in Sudan: Lessons for the EU common foreign and security policy. research papers. Retrieved from
  28. Fouéré, E. (2016). The EU special representatives: A dwindling but resilient resource at the service of EU foreign and security policy. 348. CEPS policy brief. CEPS: Brussels. Retrieved from
  29. Gastinger, M. (2017). Effects of contestation within a collective agent in EU trade policy-making. In T. Delreux & J. Adriaensen (Eds.), The principal–agent model and the European Union (pp. 181–202). London: Palgrave MacMillan.Google Scholar
  30. Girke, N. (2015). A matter of balance: The European Union as a mediator in Yemen. European Security, 24(4), 509–524.CrossRefGoogle Scholar
  31. Grevi, G. (2007). Pioneering foreign policy: The EU special representatives. Paris: European Union Institute for Security Studies.Google Scholar
  32. Haftel, Y., & Thompson, A. (2006). The independence of international organizations: Concept and applications. The Journal of Conflict Resolution, 50(2), 253–275.CrossRefGoogle Scholar
  33. Hawkins, D., & Jacoby, W. (2006). Delegation under anarchy: States, international organizations, and principal–agent theory. In D. Hawkins, D. Lake, D. Nielson, & M. Tierney (Eds.), Delegation and agency in international organizations (pp. 199–228). New York: Cambridge University Press.CrossRefGoogle Scholar
  34. Hawkins, D., Lake, D., Nielson, D., & Tierney, M. (Eds.). (2006). Delegation and agency in international organizations. New York: Cambridge University Press.Google Scholar
  35. Helly, D., Herrero, A., Knoll, A., Galeazzi, G., & Sherrriff, A. (2014). A closer look into EU’s external action frontline framing the challenges ahead for EU delegations. 62. Ecdpm Briefing Note. European centre for development policy management: Maastricht. Retrieved from
  36. Helwig, N. (2017). Agent interaction as a source of discretion for the EU high representative. In T. Delreux & J. Adriaensen (Eds.), The principal–agent model and the European Union (pp. 105–129). London: Palgrave MacMillan.Google Scholar
  37. Helwig, N., Ivan, P., & Kostanyan, H. (2013). The new EU foreign policy architecture: Reviewing the first two years of the EEAS. Brussels: Centre for European Policy Studies (CEPS).Google Scholar
  38. Hill, C. (2001). The EU’s capacity for conflict prevention. European Foreign Affairs Review, 6(3), 315–333.CrossRefGoogle Scholar
  39. Howorth, J. (2010). The EU as a Global Actor: Grand Strategy for a Global Grand Bargain. Journal of Common Market Studies, 48(3), 455–474.Google Scholar
  40. Kiewiet, R., & McCubbins, M. (1991). The logic of delegation: Congressional parties and the appropriations process. Chicago: University of Chicago Press.Google Scholar
  41. Klein, N. (2010). European agents out of control? Delegation and agency in the civil-military crisis management of the European Union 1999–2008. Baden-Baden: Nomos publisher.Google Scholar
  42. Laloux, T. (2017). Designing a collective agent for trilogues in the European parliament. In T. Delreux & J. Adriaensen (Eds.), The principal–agent model and the European Union (pp. 83–103). London: Palgrave MacMillan.Google Scholar
  43. Leblond, P. (2011). EU, US and international accounting standards: A delicate balancing act in governing global finance. Journal of European Public Policy, 18(3), 443–461.Google Scholar
  44. Lederach, J. (1995). Preparing for peace: Conflict transformation across cultures. Syracuse: Syracuse University Press.Google Scholar
  45. Marangoni, A., & Vanhoonacker, S. (2015). The consistency of EU external action post-Lisbon: Renewing appropriate measures against Zimbabwe, in 2012. European Integration Online Papers, 19(1), 1–22.Google Scholar
  46. Miall, H. (2004). Conflict transformation: A multi-dimensional task. In A. Austin, M. Fischer, & N. Ropers (Eds.), Transforming ethnopolitical conflict: The Berghof handbook (pp. 1–18). Wiesbaden: VS Verlag für Sozialwissenschaften.Google Scholar
  47. Morillas, P. (2015). From policies to politics: The European Union as an international mediator in the Mediterranean. 23. PapersIEMed/EuroMeSCo. Retrieved from
  48. Motsamai, D., & Brosig, M. (2014). Modeling cooperative peacekeeping: Exchange theory and the African peace and security regime. Journal of International Peacekeeping, 18(1–2), 45–68.Google Scholar
  49. Niemann, A., & Huigens, J. (2011). The European Union’s role in the G8: A principal–agent perspective. Journal of European Public Policy, 18(3), 420–442.CrossRefGoogle Scholar
  50. Pinfari, M. (2013). The EU, Egypt and Morsi’s rise and fall: “Strategic Patience” and its discontents. Mediterranean Politics, 18(3), 460–466.CrossRefGoogle Scholar
  51. Pirozzi, N. (2013). The EU´s comprehensive approach to crisis management. EU crisis management papers series. DCAF: Brussels. Retrieved from
  52. Plank, F. (2017). The effectiveness of interregional security cooperation: Evaluating the joint engagement of the EU and the AU in response to the 2013 crisis in the Central African Republic. European Security. doi:  10.1080/09662839.2017.1327849.
  53. Pollack, M. (2003). The engines of European integration: Delegation, agency, and agenda setting in the EU. Oxford: Oxford University Press.CrossRefGoogle Scholar
  54. Rein, C. (2015). The EU and peacekeeping in Africa: The case of AMISOM. Global Affairs, 1(2), 193–204.CrossRefGoogle Scholar
  55. Smith, K. (1998). The use of political conditionality in the EU’s relations with third countries. How effective? European Foreign Affairs Review, 3(1), 253–274.Google Scholar
  56. Tercovich, G. (2014). Towards a comprehensive approach: The EEAS crisis response system. Journal of Contingencies and Crisis Management, 22(3), 150–157.Google Scholar
  57. Tocci, N. (2007). The EU and conflict resolution: Promoting peace in the backyard. New York: Routledge.Google Scholar
  58. Tolksdorf, D. (2014). EU diplomacy at the individual level: The role and impact of the EU special representatives. In J. Koops & G. Macaj (Eds.), The EU as a diplomatic actor (pp. 69–87). Palgrave: Basingstoke.Google Scholar
  59. Vanhoonacker, S., & Pomorska, K. (2013). The European external action service and agenda-setting in European foreign policy. Journal of European Public Policy, 20(9), 13161331.CrossRefGoogle Scholar
  60. Welz, M. (2016). Multi-actor peace operations and inter-organizational relations: Insights from the Central African Republic. International Peacekeeping, 23(4), 568–591.CrossRefGoogle Scholar
  61. Zürn, M. (2012). Global governance as multi-level governance. In D. Levi-Faur (Ed.), The Oxford handbook of governance (pp. 80–99). Oxford: Oxford University Press.Google Scholar

Copyright information

© The Author(s) 2017

Authors and Affiliations

  • Friedrich Plank
    • 1
  • Arne Niemann
    • 1
  1. 1.Johannes Gutenberg University of MainzMainzGermany

Personalised recommendations