Unintended Consequences of EU External Action

ABSTRACT

There is a gap in IR and EU scholarship concerning unintended consequences in an international context, leaving this important phenomenon understudied. To fill this gap, a conceptualisation of unintended consequences is offered, and a set of common research questions are presented, highlighting the nature (what), the causes (why) and the modes of management (how) of unintended consequences of EU external action. The Special Issue contributes to the study of the EU as an international actor by broadening the notion of the EU’s impact abroad to include the unintended consequences of EU (in)actions and by shedding new light on the conceptual paradigms that explain EU external action.

KEYWORDS:

• unintended consequences
• international relations
• EU external action; foreign policy

In international relations (IR), unintended consequences are abundant, but mostly go unstudied. Although various IR theories allow for unintended consequences in one way or another, systematic explorations of unintended consequences as such in IR and international studies more generally are rare. Edited publications on unintended consequences of peacekeeping operations (Aoi et al. 2007), international security governance (Daase and Friesendorf 2010), migration law and policy (Dickie et al. 2016) and international development cooperation (Koch and Schulpen 2018) are only exceptions.

Meanwhile, the European Union (EU) has emerged as an ambitious and active global force. The studies of EU external action, however, are still mostly focused on its sui generis character (and so somewhat detached from broader IR scholarship) and much less on the consequences of its actions. The mainstream literature frames EU external performance in terms of success or failure to achieve the intended effects (usually defined on the basis of the EU’s own declared objectives). Resting on a deeply embedded and uncritical liberal assumption that EU engagement abroad is a good thing, these analyses are framed as ‘positive impact or no impact’, reducing the study of EU impact to what is essentially a study of EU effectiveness and neglecting frequent, multiple and varied unintended effects of EU policies (Burlyuk 2017, 1009-1010). The empirical reality, however, suggests that EU external action often changes matters in unintended ways, pushing the notion of unintended consequences to the forefront and begging for a revision of our models for analysing and assessing EU impact.

The lack of research on unintended consequences is striking. While unintended effects have been addressed occasionally in the context of the EU’s internal policies (e.g. Allwood et al. 2013; Dimitrakopoulos 2001; Murphy 2013; Pierson 1996; Reichert 2010) and external policies (Boerzel and Pamuk 2012; Martens and Wolf 2009; Stevens 2006), no comprehensive studies have been conducted so far. Moreover, when not omitted altogether, the concept of unintended consequences is used by IR and EU scholars casually, without much methodological or conceptual rigour. While the term ‘unintended consequences’ is self-explanatory and refers to outcomes of purposive action(s) that are not directly intended by an actor, its standard meaning today has been reduced to unwelcome unanticipated policy outcomes – an extremely partial understanding of the concept reflecting but one of its possible variations. In an earlier publication, Olga Burlyuk (co-editor of this issue) makes a plea for (re)introducing the notion of unintended consequences into our understanding of impact abroad and putting it firmly, and indeed urgently, on the scholarly research agenda (Burlyuk 2017).

This Special Issue seeks to help close these three clear analytical gaps. The objective is thus to offer a systematic analysis of the concept of unintended consequences in international relations using the case of EU external action. Owing to its hybrid nature, combining features of an international organisation and a state, the EU offers a particularly interesting and fruitful setting for examining the multiple manifestations of unintended consequences in an international context, making the findings of this issue relevant to scholars and policymakers of international organisations, states and the EU itself.

To address the gaps in knowledge identified above, we have formulated three guiding questions for this Special Issue:

• What are the diverse manifestations of unintended consequences in various functional domains of EU external policies?

• Why do these unintended consequences occur?

• How are these unintended consequences managed by the actor concerned in the international context?

The remainder of this Introduction is organised as follows: after a discussion of the baseline conceptualisation of unintended consequences from which this collective investigation departed, and the ‘sampling for diversity’ of the contributions, we elaborate on the findings and the contribution the issue makes to the study of unintended consequences (in international relations) and the EU as an international actor.

What’s in a name: a baseline conceptualisation of unintended consequences

To approach the ‘what’ question formulated above, this project departed from a shared baseline conceptualisation of unintended consequences as found in the literature on unintended consequences in sociology and beyond, and as synthesized for the purposes of IR and EU studies by Burlyuk (2017). Fundamentally, this conceptualisation challenges the standard understanding of unintended consequences as unwelcome unanticipated effects which, we argue, reflects only one of the concept’s possible variations.

At the most general level, the concept-and-term unintended consequences refers to outcomes of purposive action(s) that are not directly intended by an actor. As argued elsewhere, when unpacked, unintended consequences may vary with respect to (at least) nine parameters, including: mode of knowledgeability, relationship to the initial intention, value attached, type of action, strength of causal links, type of outcome, who is impacted by it, mode of acknowledgement and, finally, the temporal dimension (for a detailed discussion see Burlyuk 2017, 1011-1017). Several of these analytical categories fall victim to their seemingly self-evident nature, and their casual use breeds conceptual conflation. It is the aim of this Special Issue to advance conceptual clarity with regard to unintended consequences. Below we focus on the dimensions that appear to be most challenging analytically or methodologically.

First of all, establishing intention so as to label something unintended is admittedly a challenging task, because an action’s purpose is not always clear-cut or explicit and may be continuously reassessed (Baert 1991, 202; Merton 1936, 896). Moreover, it is often more appropriate to speak of multiple or composite intentions guiding one’s action. Intention, or the actual purpose of a given action, “is, in practice, almost impossible to ascribe” (Buckland 2010, 138). A number of proxies for intention can be used, actor’s stated objectives being the most obvious and readily available one. Needless to say, however, when studying a political actor in an international context, one has to be wary not to take political declarations and policy documents at face value. They have to be problematized, for example, by clarifying stated intentions, probing hidden intentions in interviews, or unpacking ‘collective intention’ into constitutive (and possibly conflicting) intents of the respective sub-actors (a particularly pertinent exercise in the case of a complex actor like the EU). Interviewing and process-tracing may thus be instrumental in studying unintended consequences. Importantly, taking stated objectives as a reference for intention so as to delineate which of the consequences of one’s action were intended and which were not is principally different from taking stated objectives as a reference for effects so as to delineate one’s impact. The latter approach ignores all effects of one’s action that do not fulfil stated objectives and thus reduces the study of impact to the study of effectiveness, neglecting large parts of empirical reality.

The most important dimension – and the one for which there is clearly the most misunderstanding and conflation with regard to unintended consequences – is the mode of knowledgeability. ‘Unintended’ in unintended consequences stands for not intended, that is: not within the purpose for which an action was undertaken. Unintended should thus not be confused with unanticipated or unforeseen. While an unintended consequence can be unanticipated in instances when an actor did not foresee it at all (which is the standard understanding and most frequent usage), it can also be anticipated in instances when an actor foresees the possibility of this outcome but follows through with the initial action and intention nonetheless (which is arguably the variation of unintended consequences most frequently observed in politics and international relations, and prominent in the contributions to this issue). The key message here is that unintended consequences are not by definition unknown to the actor undertaking the action (Baert 1991, 201; Burlyuk 2017, 1013). Consequences that are both unintended and anticipated are a real category of the phenomenon, and arguably the most interesting and least studied in IR and EU studies.

Similarly, when it comes to the value attached to an unintended consequence by the actor undertaking the action, the target of the action or a third party, it is important to remember that unintended consequences are not axiologically negative: “undesired effects are not always undesirable effects” (Merton 1936, 895). An unintended outcome may be desirable (beneficial), undesirable (detrimental) or neutral, depending on the value attached to it by the actor and others (Baert 1991, 204; Burlyuk 2017, 1013; Kohstall 2010, 10). Although negative unintended consequences are more in the public eye and may rightfully be the more intriguing analytical variation to study, it is important to remember that positive unintended consequences may seem less interesting but often prove to be no less significant.

The value attached to an unintended consequence (positive/negative) together with the mode of knowledgeability about it (anticipated/unanticipated), when combined, produce four variations of unintended consequences: positive anticipated, positive unanticipated, negative anticipated and negative unanticipated.¹ As noted earlier, the term unintended consequences is almost exclusively associated with this last type: negative unanticipated. Yet it is just one of the four possible variations of the concept. In fact, it is unintended consequences which are both negative and anticipated that are particularly interesting to study, because they make one engage with the ethical dimension of policymaking and address the dilemmas an actor faces in this regard. As problematized by de Zwart (2015, 294), since anticipated unintended consequences often concern politically sensitive and otherwise controversial issues, “the connotation ‘unanticipated’ that sticks to ‘unintended’ makes it possible to expose, discuss, or correct such issues without imputing blame or getting into painful discussions about responsibility”. De Zwart concludes that “the distinction between intention and anticipation is clearly a question in ethics”: why, given the foreseen risks, do policymakers choose to go ahead anyway (295)? Likewise, positive unintended consequences, both anticipated and unanticipated, are mostly ignored in policy discussions, yet policies can sometimes bring bonuses and windfalls that were not factored into the initial policy designs.

Another analytical clarification in order here relates to the relationship of an unintended consequence to the initial intention. While the term unintended consequences is associated commonly with outcomes that frustrate one’s intention, an unintended consequence may just as well fulfil the intention behind an action or have altogether no effect thereon (Baert 1991, 24; Burlyuk 2017, 1014; Kohstall 2010, 10). For example, one can have an unintended consequence that is deemed negative by the actor and yet fulfils the actor’s initial intention. An important implication of this consideration is that having unintended consequences is analytically distinct from failing to achieve intended consequences, and thus the two must not be conflated. Failure to reach intended consequences may well entail some unintended ones, just as unintended consequences may frustrate the initial intention and so lead to failure to achieve intended effects, but not inevitably so. Unintended consequences may reinforce the action and so facilitate the achievement of intended consequences, or they may backfire or lead to consequences that are altogether unrelated to the initial intention. Similarly, the study of unintended consequences must not be conflated with the study of policy failure or the usual policy critique.

Just as unintended consequences are not necessarily unanticipated or undesirable, they are not necessarily consequences for the actor undertaking the action (as is generally implied). With regard to who is impacted, one can distinguish between unintended consequences that affect the actor itself, those who are targeted by the action and third parties (Burlyuk 2017, 1015; Merton 1936, 895). One can thus speak of internal unintended consequences (with impact on the actor undertaking the action) and external unintended consequences (with impact on those targeted by the action or third parties). This internal/external consideration is relevant for identifying unintended consequences in the first place, as well as for analysing how policymakers prepare themselves for consequences they anticipate, and react to those they do not. In addition to such ‘individual effects’, unintended consequences can also produce systemic or aggregate effects at the level of the social (in our case, political, international) system (Baert 1991, 203-204).

Finally, unintended consequences may have stronger or weaker causal links to the action and the actor (Jervis 1997/1998; Schneckener 2010). This is not to suggest that causal links are unimportant: on the contrary, it is crucial to maintain a clear distinction between consequences (as effects that are causally connected to the policy action under study) and external shocks (as events that may have significant impact on the outcome under study but are not a consequence of the policy action under consideration). To put it simply, consequences are to be distinguished from comets.² Causal mechanisms may be vague, ambiguous and subject to narration, so it is the responsibility of the scholar to present a compelling story of the causal link between the action and what is claimed to be and is studied as its consequence.

The fact that analysing unintended consequences comes with a set of methodological challenges and is often regarded as “a troublesome undertaking” (Zingerle 1998, 184) does not justify abandoning the task altogether. On the contrary, it means that scholars have to be extra rigorous. In fact, the difficulties in ascribing intent, attributing intent to action, linking action to actor and outcome to action, which are all too often mentioned as reasons to dismiss the idea of analysing unintended consequences, are just as pertinent to the study of intended consequences as they are to the study of the unintended. Just as some 80 years ago, when Robert Merton wrote these words, the puzzle “must be solved for every empirical case which is studied” (Merton 1936, 897).

As for the causes behind unintended consequences, or the ‘why’ question asked in this Special Issue, the unintended consequences literature rests on assumptions about the world’s inherent complexity and inevitable cognitive and emotional biases and, accordingly, attributes unintended consequences to ignorance, error, imperious immediacy of interests, basic values, self-defeating predictions (Merton 1936; 1948), as well as willingness to risk, indifference (to possible future harm) and contextual change (Almondo 1998; Burlyuk 2017, 1015-1017; de Zwart 2015, 295; Vernon 1979, 68). Authors in this issue use these basic explanatory factors as a starting point and further explore literature in the broader field of IR to give a more solid conceptual grounding to the explanation of their specific cases.

As for the modes of managing unintended consequences, or the ‘how’ dimension of this study, the unintended consequences literature has done little so far to systematise the ways in which actors respond to the unintended consequences of their actions or act in anticipation of them. By engaging with the how question throughout the issue, and by asking the how question alongside the what and the why questions, this Special Issue makes an original contribution to the study of unintended consequences (as such, and IR/EU studies in particular) and devises a classification of the modes of managing unintended consequences that can be used in further research on the subject (more on this below).

What’s in the Special Issue: sampling for diversity

In soliciting contributions for this issue, we sampled for diversity with regard to analytical variations of unintended consequences under study, functional domains of EU external action and targeted geographic areas (partner states, groups of states or regions).

With respect to the variation in the types of unintended consequences, contributions in this Special Issue explore positive and negative unintended consequences, anticipated and unanticipated unintended consequences, unintended consequences stemming from the EU’s action and inaction (recognising failure to act as a form of action), and unintended consequences that affect the EU, the targets of its policies and third parties. Some articles identify and analyse unintended consequences of an EU policy in a specific domain at a more general level (De Ville and Gheyle; Reslow; Kourtelis; and Dandashly and Noutcheva³ ), while others unpack unintended consequences in concrete cases (Carbone; Casier; Bouris). Some examine unintended consequences that have happened (De Ville and Gheyle; Reslow; Bouris; Dandashly and Noutcheva). Some study the unintended consequences of the policies (Reslow; Kourtelis; Bouris; Dandashly and Noutcheva) while others study policies as unintended consequences (Carbone; Casier).

In terms of policy areas, articles examine instances of unintended consequences of EU neighbourhood and enlargement policies, specifically with regard to democratisation (Dandashly and Noutcheva) and state-building (Bouris), EU external (rural) development (Carbone; Kourtelis), trade (De Ville and Gheyle), migration (Reslow), EU diplomacy and negotiations with (groups of) third states (De Ville and Gheyle; Carbone), CSDP missions (Bouris) and foreign policy as such (Casier).

The geography of contributions stretches from EU-US (De Ville and Gheyle) and EU-Russia (Casier) relations, through EU relations with the eastern and southern neighbourhoods (Dandashly and Noutcheva), and specifically the Arab Mediterranean (Kourtelis) and Israel-Palestine (Bouris) to Africa (Reslow), in particular the African, Caribbean and Pacific Group of States (Carbone).

Such analytical and empirical breadth and depth of the topics guarantees a wide appeal for IR and EU scholars and policymakers and strengthens the study’s comparative dimension, even though it was not designed for the purpose of comparison as such. While authors were free to choose the theoretical and methodological lenses best suited to their studies, all contributions are united by the empirical phenomenon in focus (unintended consequences of EU external action) and the three guiding research questions (the what, why and how). Furthermore, they all departed from the shared baseline understanding of unintended consequences discussed above. The findings of this Special Issue thus make important contributions to the literature on unintended consequences in general and unintended consequences in IR/EU external action in particular and to the literature on the EU as an international actor (summarised below), as well as to (sub)literature on the policy areas and/or cases examined (discussed in individual articles).

Contribution to the study of unintended consequences

First, the contributions in this Special Issue reveal that unintended consequences can result not only from purposive action, as the baseline conceptualisation of unintended consequences discussed above put it (what), but also from purposive inaction (Casier; Bouris; Dandashly and Noutcheva). Purposiveness, intentionality, awareness of one’s act (be it action or inaction), the fact that it “involves motives and consequently a choice between various alternatives” (Merton 1936, 895), “the want” at the moment of carrying out the act (Baert 1991, 201), are the decisive criteria. The study of (international actors’/the EU’s) inaction, indecision, non-policy thus emerges as a field for further investigation, adding an entirely new dimension of empirical reality to the research agenda. At the very least, it pushes scholars to engage in policy analysis that goes beyond policy decisions as recorded in policy documents.

Second, in their own way all contributions to this Special Issue advance an understanding of unintended consequences as a process: not only can policies generate unintended consequences (Casier; Bouris; Dandashly and Noutcheva), but existing or anticipated unintended consequences can also drive or generate policies (Reslow; Carbone; Kourtelis). Unintended consequences can even trigger further unintended consequences (De Ville and Gheyle; Casier). A notion of first-order and second-order unintended consequences is proposed in this regard to capture and reflect the causal and temporal distance between one’s (in)action and the unintended effects it triggers (De Ville and Gheyle). This begs for non-linear scholarly accounts of (EU) policies and their effects that view the two as parts of one dynamic process rather than strictly sequential steps. It also invites analyses that consider interactions at all levels: across (sub)actors, for example, between the EU, its partners and third states (Casier; Dandashly and Noutcheva) or within the EU itself (Reslow; Carbone); across policy areas, acknowledging and exploring linkages between different functional domains (Reslow; Carbone) and between internal and external policy levels (Reslow); across time, with learning from the past (Kourtelis); and, finally, across space, with learning from others and others learning from you.

Third, contributions in this Special Issue suggest that the unintended consequences of policies (and specifically international policies) are often or, at least, can often be anticipated (De Ville and Gheyle; Reslow; Kourtelis). This entails that unintended consequences are not only and not so much about knowing or not knowing the possible effects of one’s (in)action as they are about one’s political and ethical choices. In some instances, policy decisions are clearly a result of weighing off against each other the different sets of consequences an actor anticipates (Reslow). With regard to unintended consequences that are both anticipated and desirable (positive), the question then becomes: at what point do these cease being unintended and become intended (Reslow)? In turn, the question with regard to unintended consequences that are undesirable (negative) is whether and why these have been “foreseeable but not foreseen, or foreseen but neglected or discounted or risked” (Vernon 1979, 61). The matters of the politics and ethics of policies – and their intended as well as unintended consequences – are in urgent need of further exploration in EU studies and international studies more generally.

Fourth, with respect to the causes of unintended consequences (why), contributions in this issue trace the drivers of unintended consequence to a plurality of factors. This Special Issue departs from the explanatory variables suggested by sociological scholarship (such as error, ignorance, interests, values, self-fulfilling prophecy, risk, indifference and contextual change), ‘domesticates’ these insights in the discipline of political science and anchors them more firmly in conceptual paradigms established in the IR/EU fields (more on this in the following section). While some analyses engage with the above categories comprehensively, entertaining all of them as possible explanations (De Ville and Gheyle; Kourtelis; Bouris), others do so in a more selective or adapted manner as required by their research objectives (Reslow; Carbone; Casier; Dandashly and Noutcheva). Notably, in one way or another all contributions support the assertion of unintended consequences literature that a plurality of causes are at work in each given case and consider this (highly) applicable to an international relations context. In this way, we acknowledge the complexity of social phenomena and join the chorus of voices that have embraced analytical eclecticism as a method of inquiry in world politics (Cornut 2014; Sil and Katzenstein 2010).

Finally, this Special Issue is among the first scholarly efforts to engage systematically with the question of how actors manage the unintended consequences their policies have caused or may cause. The contributions reveal that the mode of managing unintended consequences may vary depending on the policy context. It is fair to say that the management of unintended consequences requires a degree of acknowledgement that such effects occur or may occur in the future. We have identified two important markers of the variety of management strategies that actors resort to. When faced with unintended consequences or the prospect of unintended consequences of policies, policymakers have to decide, first, what value they attach to them. In other words, the policy response is conditioned by the judgement made with respect to whether the unintended consequences are seen as positive, negative or neutral. This is a dynamic process in which policymakers assess the magnitude of the unintended effects, the party experiencing the unintended effects and the possibility of shifting perceptions of policy effects in time. This decision, in turn, conditions whether policymakers wish to see these unintended effects mitigated or reinforced.

A second important aspect of decision-making involves choosing whether to change policy so as to address its unintended effects or continue with the existing policy and thus embrace those unintended effects. The decision to act or not to act is also complex in nature and depends on the desirability of the unintended policy outcomes, the costs of action or inaction, and the ethics of prioritising one course of action over another, among other situational factors. The judgement about these parameters will inevitably evolve over time alongside policymakers’ willingness and ability to change the policy or not.

Contribution to the study of the EU as an international actor

Multiple manifestations

Unintended consequences in IR and EU external action are ubiquitous. Our Special Issue supports the initial hunch behind this collective investigation with evidence from a variety of EU policy areas and bilateral and multilateral relationships. The fact that the authors have (rather easily) found examples of unintended consequences in their respective fields of expertise once again confirms that unintended consequences are there, are many and diverse, and are waiting to be examined. The need to integrate unintended consequences in our scholarly and policy notions of ‘impact’ and respective research/policy agendas is thus truly pressing.

Not only are unintended consequences of EU external action abundant but they are also diverse in nature. An important finding of this issue is that most of the unintended consequences of EU external (in)action uncovered by the contributors were anticipated. The fact that unintended consequences occur despite their predictability raises further questions about the reasons for allowing them to manifest themselves. While the majority of unintended consequences addressed in this Special Issue were negative, quite a few have a positive value attached to them from the point of view of the EU as a whole (De Ville and Gheyle), different institutional actors within the EU (Reslow) or third parties outside the EU (Reslow; Carbone).

Another significant finding is that the EU is likely to recognise the unintended consequences of its policies and this, in theory, lays the ground for active management of them. Furthermore, unintended consequences that affect actors other than the EU or the system as such – described as ‘external effects’ in some contributions (Reslow; Casier; Dandashly and Noutcheva) – are encountered far more frequently than ‘internal effects’, which raises the importance of the ethical dimension of EU external (in)action. In any case, the contributions consider not only the external effects of EU external policies (Kourtelis; Bouris; Dandashly and Noutcheva), but also the internal effects of EU external policies (Reslow; Carbone; Casier) as well as the external effects of EU internal policies (De Ville and Gheyle; Reslow).

Multiple causes

Unintended consequences have not only multiple manifestations but also multiple causes. We contribute to the literature on the EU’s international role by opening up the conceptual debate about the sources of EU external action and by going beyond the unproductive dichotomy of interests versus values which has dominated explanatory accounts of the EU’s external actorness till now. Some have insisted on the traditional state-like features of the EU and attributed its external actions to a conventional conception of interests, portraying it in either realist (Hyde-Price 2006) or liberal (Smith 2011) terms. Others have gone beyond the rationalist paradigm and highlighted the normative fundamentals of the EU’s external posture, conceptualising the EU as a normative power and debating the extent to which the EU can make a normative difference in world affairs (Diez 2013; Manners 2002; Sjursen 2006; Whitman 2013). Inherent in this debate is the discussion about whether the EU is an actor like any other or a sui generis entity that is bound to have a unique imprint on the global scene.

While we recognise the importance of these fundamental motivations of external policy, we focus on less studied drivers of external action reaching out to literature outside the EU mainstream. Although authors were free to engage with theories of their choice, the contributions in this Special Issue fall within one (or more) of three perspectives that have proven fruitful in providing alternative accounts of unintended effects of external action: risk, bureaucracy and policy learning. As these come from broader non-EU specific literature, they can be applied to the external actions of various players in an international context.

Firstly, the literature on risk has informed our understanding of how risk judgements interfere in establishing hierarchies of anticipated unintended consequences and how ethics form an integral part of the decisions to tolerate anticipated undesired effects or not (De Ville and Gheyle; Kourtelis; Bouris). This is a fundamental insight that moves the scholarly debate beyond rational choice paradigms and into the realm of scholarship emphasizing risk and how risk shapes decisions, policies and society more broadly (Beck 1992). While mainstream rational approaches conceptualise decision-making as a cost-benefit calculation with preconceived notions of what constitutes a gain or a loss, risk-based accounts capture the complexity of decision-making when trade-offs between various courses of actions exist and when those policy alternatives carry different risks of undesired effects (Webb 2018). Decisions are not simply reasoned in terms of gains and losses, but involve an interaction between the risk assessments of policy outcomes, the ethical positions of policymakers and the anticipation of undesirable unintended consequences (De Ville and Gheyle; Kourtelis; Bouris).

The EU is in this sense presented as an entity making conscious choices about the unintended consequences of policies it will tolerate, which, in turn, have serious ethical implications for its international reputation (De Zwart 2015). This sheds new light on debates about the EU’s ethical power which have focused exclusively on the extent to which the EU’s actions are self-regarding or take the interests and values of the parties receiving EU policy into account (Aggestam 2008). A risk perspective suggests that the ethics of EU power may have to do with the weight the EU places on different foreign policy goals and how it judges the desirability of those goals in time and their potential to trigger different types of unintended consequences for different actors.

Secondly, the literature on the ‘pathologies of international organisations’ provides another alternative prism through which unintended consequences are viewed as the outcome of dysfunctional bureaucratic behaviour (Barnett and Finnemore 2004). As a dense institutional environment, the EU may be particularly prone to errors and misjudgements linked to the very nature of its bureaucracy. The obsession with internal rules and procedures, as well as the compartmentalisation of expertise across various institutional units, increases the danger of misreading and responding inadequately to the global and local contexts in which external action is taken (Reslow; Casier; Dandashly and Noutcheva). In spite of its highly professional bureaucracy, the EU is not unaccustomed to artificially copy-pasting templates, policy tools and solutions from one context to another, counting on previous successes while neglecting the specificity of new environments (Kourtelis).

Importantly, the administrative governance of overtly politicised dossiers with geostrategic implications may carry particular risks of unintended consequences in an international context (Dandashly and Noutcheva). Technocratic approaches are well known to rely on an internal logic of policymaking that may ignore the intensity of preferences of third actors and fail to anticipate their reactions. The dysfunctional bureaucratic logic thus brings in a new perspective on the EU foreign policy bureaucracy which has mostly been examined through the positive input of its unelected officials and their subtle yet influential role in setting the EU’s foreign policy agenda (Vanhoonacker and Pomorska 2013). The bureaucratic propensity for self-defeating behaviour suggests that unintended consequences are inherent in the way in which the Brussels bureaucracy functions and produces policy output with external implications.

A particular variant of this perspective looks at the role of politics in decisions related to external action and conceives of unintended consequences as the outcome of internal dynamics that juxtapose different foreign policy objectives within and across institutions and within and across different layers of governance. This is especially relevant for the EU which is a complex decision-making system in which intergovernmental and supranational elements of policy-making intersect at different moments of the policy cycle, in addition to the usual challenges of cross-sectoral policy coordination of every state administration. EU scholars have discussed at length the problem of coherence in EU foreign policymaking and how it often leads to the EU’s suboptimal external performance (Carbone 2013; Gebhard 2017).

We go a step further by showing how the competing objectives of different policy sectors, as well as the propensity of EU internal policies to have external effects, can produce unintended consequences for different stakeholders in this complex policymaking system (Reslow; Carbone; Dandashly and Noutcheva). We find that this cacophony of relationships between member states and the EU, between supranational and intergovernmental modes of governance, between inter- and intra-institutional interactions within each domain of external action is a powerful source of unintended consequences of EU external action. Various contributions provide detailed evidence of how they play out in different policy domains.

Thirdly, the literature on policy learning offers another analytical angle through which unintended consequences can be studied (Bennett and Howlett 1992). In principle, the EU can learn from its past mistakes by adapting its own policy script, but the process of acknowledging the unintended effects of one’s policies is fundamentally challenging (Kourtelis; Casier; Bouris). The main insight of the scholarship on policy learning relevant for the discussion here is that institutional actors are able to acknowledge and anticipate the unintended consequences of policy initiatives and can use this knowledge strategically to sway future policies in a preferred direction (De Ville and Gheyle; Carbone; Biedenkopf 2018; Koch 2018). Politics is at the core of this process, where advocacy coalitions within or outside the EU institutional environment can instrumentalise a narrative on unintended consequences in order to justify a policy change or initiate a new policy on the grounds that it reduces the likelihood of future unintended consequences. Unintended consequences can thus be politicised and used strategically by actors in their internal institutional and policy battles (De Ville and Gheyle; Carbone). A policy learning perspective thus offers new insights into how the occurrence of unintended consequences can in itself become the cause of policy change, bringing in a transgovernmental aspect of policymaking and emphasising the importance of agency in constructing a narrative on unintended consequences that can be harnessed for the purposes of policy entrepreneurship.

These three perspectives are not exhaustive in terms of conceptual accounts of unintended consequences. For example, none of the contributions to this Special Issue borrow insights from the various approaches to foreign policymaking grounded in political psychology, such as groupthink or cybernetic decision-making, which could potentially provide points of departure for further conceptual exploration. Nevertheless, the three perspectives emphasized by the contributions, combined with a sensitivity to the plurality of causes at work discussed in the previous section, make it possible to overcome the unproductive dichotomies of norms versus interests, values versus security, etc. that have characterised much of the mainstream EU scholarship. The Special Issue thereby contributes to a growing body of research that engages pragmatically with conceptual paradigms to construct complex causal narratives while remaining closer to the problems policymakers face on a daily basis.

Multiple strategies

There are various ways in which actors can manage the unintended consequences of their policies, and this Special Issue makes a particular contribution to this subject not addressed systematically in previous scholarship. Our study uncovers considerable variations in how the EU manages unintended consequences, allowing us to draw parallels across policy sectors and partners; although a rigorous comparison and explanation of the variations observed is beyond the scope of this issue. What determines the exact choice of management strategy is an interesting question but one that needs to be addressed in future research. Our study, a first attempt at classifying EU responses to the unintended consequences of its policies, generates far more interesting questions that it can answer. It does, however, suggest a few avenues for further exploration.

The contributions to the Special Issue find that the EU is at times a reactive and at times a pro-active manager of unintended consequences. In some cases where the EU was able to anticipate undesirable effects of default policies, it adopted a pro-active approach, such as in the negotiations of a new legally binding EU-ACP partnership (Carbone). In other cases where undesirable unintended consequences could have been anticipated but the EU failed to connect the dots, it chose to be inactive and stick to its ‘business as usual’ approach, such as in its policies vis-à-vis the neighbouring countries (Kourtelis; Dandashly and Noutcheva). There have also been instances in which the EU has produced unintended consequences for third parties (Casier). In such cases, management of the unintended effects of EU policies has also been influenced by the agency of the third actors concerned and their reactions to the EU’s policies (Casier).

A specific marker of the EU’s approach to management of unintended consequences is the strength of its legal competences and institutional mandate in specific domains of external action. In the area of trade policy, for example, in which the EU has exclusive competence and enjoys a very strong institutional mandate, it has shown itself to be a daring actor, initiating substantive policy change in the aftermath of the failure of the TTIP negotiations (De Ville and Gheyle). This has not been the case in the migration policy field, where the fragmented nature of decision-making and the constrained mandate of the EU institutional layer make it impossible for the EU to show strong leadership in tackling unintended consequences (Reslow). A similar conclusion emerges in the area of democracy promotion, where the EU institutional establishment is not in a position to drive substantive policy change without the consensual backing of EU member states (Dandashly and Noutcheva). Furthermore, in the area of state-building, where sensitivities are very high and EU member states are in the driver’s seat, the EU is in a very weak position and reluctant to admit that its interventions can have unintended consequences, let alone actively manage them (Bouris).

In choosing what management strategy to adopt, at times the EU also faces ethical dilemmas that involve weighing off the different unintended consequences and the effects they have on the EU and the other parties (Reslow). Aware that it is impossible to eliminate unintended consequences as a category and that it is a political actor’s job to anticipate (so that, in reality, most unintended consequences are anticipated or at least anticipatable), this Special Issue builds up to a question for further research, namely, the politics and ethics of choosing the unintended consequences that merit more (urgent) attention and active management. The concluding article by Frank de Zwart and Karolina Pomorska develops this point further and offers additional reflections on the implications of these ethical dilemmas for policymaking.

Acknowledgements

This research has been supported by a Postdoctoral Grant of the Research Foundation – Flanders (FWO), a conference grant by the Royal Netherlands Academy of Arts and Sciences (KNAW), and a research grant by the Faculty of Arts and Social Sciences of Maastricht University.

The authors are grateful to Tom Delreux, Hylke Dijkstra, Michael Emerson, Jonathan Joseph, John Peterson, Karolina Pomorska, Vsevolod Samokhvalov and all others who have provided valuable feedback on this project at the 2017 UACES conference, the 2018 EUIA conference, the 2018 Author Workshop and in the course of peer review.

Additional information

Notes on contributors

Olga Burlyuk

Olga Burlyuk is a Postdoctoral Fellow of the Research Foundation – Flanders (FWO) at Ghent University, Ghent, Belgium, and a Visiting Scholar at the Ukrainian Research Institute at Harvard University, Cambridge, US. Email: [email protected]

Gergana Noutcheva

Gergana Noutcheva is Associate Professor at the Department of Political Science at Maastricht University, the Netherlands.

Notes

¹ We are thankful to the anonymous reviewer for suggesting this taxonomy.

² We owe this analogy to the anonymous reviewer.

³ When the reference is without a date, it is to an article contained in this Special Issue.