From EU battlegroups to Rapid Deployment Capacity: learning the right lessons?

the EU’s rapid crisis-response capacity

International Affairs 100: 1, 2024 constitutes a litmus test for the success of the Strategic compass.A successful RDC would help the EU become a more credible and effective rapid crisis responder, able to act autonomously from other actors in a range of situations. 6ut has the EU accurately identified, accepted and institutionalized the right lessons from the battlegroups failure?We use the case of the development of the EU battlegroups to the RDC to better understand the capacity of the EU to learn in a way that improves its effectiveness in the domain of its military Common Security and Defence Policy (CSDP).In doing so, this article contributes not only to the literature on learning in the CSDP, but also offers insights about the specific challenges of political-military learning in multinational regional organizations.Specifically, we develop a theoretical framework to capture the most significant factors affecting learning processes by drawing on insights from the literatures on organizational learning in public policy, military innovation and learning in military organizations and CSDP in particular.We then put this into practice by studying the actors, substance and appropriateness of EU learning during three distinct periods: the initial conception of the battlegroups (2000-07); the period during which the battlegroups were on standby, after reaching full operational capability (2007-16); and the period of reform, in conjunction with the establishment of the EU RDC, which began with the adoption of the EU Global Strategy (2016-ongoing).Subsequently, we use the theoretical framework to explain and evaluate the findings, especially the role played by organizational factors and underlying learning pathologies.
Methodologically, we apply a qualitative strategy using empirical evidence gathered from official EU and member state documents, 13 elite interviews with current and former officials involved in the design and adaptation of the battlegroups, and the personal experience of one of this article's authors in the EU's military structures.We also use data from a range of unpublished non-papers and memos that were made available to us, to look beyond formal codifications of lessons learned and open up to informal knowledge transfer. 7arning processes related to rapid-reaction forces and the Common Security and Defence Policy Scholars interested in the development of EU rapid-reaction forces have focused predominantly on identifying the origins of the EU battlegroups 8 and explaining their non-use, either in general or in reference to specific crises where their deploy-International Affairs 100: 1, 2024 ment was requested or expected. 9Even as this literature recognizes several structural obstacles, such as a lack of political will, dysfunctional command and control structures and flawed funding provisions, it does not investigate the capacity of the EU to learn from these non-deployments.A similar pattern can be observed in the study of other multinational rapid-reaction forces, such as the NATO Response Force 10 or the African Standby Force, 11 where most analyses identify political obstacles and institutional design anomalies before making a range of policy recommendations towards improvement.Multinational rapid-response forces in the EU and NATO have in common that they have hardly been used so far, making them look like cases of ill-designed instruments creating programmatic failures. 12lthough each of these organizations is-or has been-confronted with the question of how to learn from failure and improve upon their rapid-reaction capacity, their ability to accurately identify and implement any lessons about the underlying causes of these difficulties has thus far not been addressed-in contrast to broader studies of learning during and after operations. 13Therefore, we lack insights into the extent to which the development of multinational rapid-reaction forces, such as the EU battlegroups, are driven by organizational learning; where such learning takes place; and how this has shaped their design, non-use and eventual redesign.
There is, however, some writing on learning in CSDP, peacebuilding and European foreign and security policy more broadly, most of which seeks to explain change through learning. 14For instance, Bossong highlighted how urgent operational pressures make learning in EU civilian crisis management possible. 15Faleg made the case that the EU's progress in the area of civilian crisis management was a result of 'learning from doing' mediated by practitioner communities, while Dijkstra, Petrov and Mahr showed how the many veto points in the EU can slow down learning in this domain. 16Most notable for the EU's military CSDP is the work International Affairs 100: 1, 2024 by Michael E. Smith, who drew on theories of organizational learning in other fields to argue that learning dynamics with feedback loops may explain the evolution of CSDP from 2003 to 2015 within a historical institutionalist framework. 17n common with other works using a learning lens in this area, Smith is neutral as to whether learning is functional or dysfunctional as a process and whether the right lessons have been identified or not.As such, his approach cannot offer specific reasons for why the EU battlegroups did not meet their original designers' expectations or why lesson-learning was so slow and ineffective for many years.
Pihs-Lang, in her unpublished PhD thesis, elaborates from the relevant literatures a theoretical framework revolving around five phases of learning to assess the first three military operations of the EU-Concordia, Artemis and Althea. 18n her findings, she outlines ten 'impact factors' that may either help or hinder learning processes in EU military crisis management. 19Like Smith, she adopts a normatively neutral approach as to whether the right lessons have been identified or learned.For both authors, learning is about a deliberate and systemic effort that can be described as successful only in so far as the learning leads to lessons being encoded in new organizational routines.This reluctance to engage with the substance of lessons is in line with much of the literature that warns against the fallacy of inferring failures of learning processes from allegedly undesirable policy outcomes.
Despite this neutral stance on the lesson substance, both Smith and Pihs-Lang advance some criticisms of the EU's learning process in security and defence.Smith refers to learning weaknesses, such as a lack of a shared learning culture across different institutions and levels within the EU's foreign policy system, and a learning gap between the civilian and military sides of peacebuilding and crisis management; insufficient progress in building a shared learning culture that extends to member states and other international partners; and limited after-event reporting or follow-up. 20Pihs-Lang refers to problems of staff rotation and insufficient handover periods, both of which can lead to institutional and case-specific knowledge being lost and not available for learning.She also notes that some lessons may be agreed, but never formally put on paper, such as the future non-viability of the Berlin Plus agreement after the Althea operation largely because of the Turkey-Cyprus issue, 21 and draws attention to the 'filtering' of lessons as they ascend the hierarchy, or the practice of having two versions of 'lessons-learned' documents-one with sensitive lessons, only for the High Representative of the EU for Foreign Affairs and Security Policy/Vice-President (HR/VP), and a sanitized version for member states.The issue of political sensitivity as a problem in the learning process is also mentioned by Bremberg and Hedling, who noted the predominant focus of officials on achieving 'quick wins' through learning rather than 'naming and shaming' member states. 22o, even though the literature on learning in military CSDP does express some critiques of learning processes, the predominant purpose is to explain the evolution of capacities and policies.Questions about whether learning is more or less likely to improve performance, through correctly identifying and tackling root causes and spreading best practices, fall largely by the wayside, both conceptually and empirically.This is regrettable in light of the extensive literature on the EU expectations/capability gap and how to narrow it.It also hinders cross-fertilization with the extensive literature on innovation in states' military affairs, most of which is interested in the link between the military's capability to innovate and its effectiveness against an adversary.23Learning attempts that miss root causes of failures, or that promote counterproductive practices, may be successful in procedural, but not in substantive terms.We need to know more about the reasons that structurally hinder the EU from engaging successfully in identifying and communicating the right lessons.In the case of the battlegroups, the specific challenge was to learn from a succession of deliberations that led to non-decisions, rather than action.

Theorizing learning in security and defence
In constructing the theoretical framework, we draw on three bodies of literature: organizational learning with specific attention to applications in military organizations;24 the evaluation of failures of learning and policy;25 and learning specifically in European security and defence. 26We share with some of the literature on military innovation and adaptation a conceptualization of learning as a process International Affairs 100: 1, 2024 and an emphasis on organizational learning capacity as a key explanatory factor.Yet, the literature's main reference point is innovation for the narrow purpose of improving operational effectiveness of states' militaries against an adversary and a focus on the interplay of changes in technology, doctrine, tactics or strategy.In our case, success needs to be related to the EU's comprehensive approach to foreign policy and crisis management and its limited defence competences.Success of learning extends to political-strategic questions, given the polity's still-evolving and at times ambiguous institutional structures, competences and instruments.The EU's Military Staff work under the political authority of a multi-hatted HR/ VP who ultimately depends on political support and agreement by member states.The participants of learning processes are thus more heterogeneous in nature, and the very purpose of learning can be contested at times.
From a broader public policy perspective, McConnell recognized the objective and subjective dimensions of success and failure.He suggested that: 'A policy fails if it does not achieve the goals that proponents set out to achieve, and opposition is great and/or support is virtually non-existent.' 27He also identified three main types of failures: process, programme and political.We argue that the case of the battlegroups is a failure at the process level with regard to lesson-learning and decision-making; a partial failure at the programmatic level, as the battlegroups did at least improve military cooperation in Europe, but failed at being useful for crisis response tasks; and predominantly a failure at the political level, as they never met their publicly articulated objectives in terms of bolstering the EU's role in security and defence. 28To explain this, we develop a three-step model.
As a first step, we develop a process-oriented understanding of organizational learning, consisting of six phases, which allows us to trace learning since the emergence of the idea of European rapid-response forces: perceptions of unusual success or failure in organizational performance, knowledge acquisition about the underlying causes through information collection and interpretation, 'identification of 'actionable lessons' to remedy or reinforce, upwards and sideways communication and diffusion of lessons, decision-making/bargaining about which lesson to adopt or adapt, and, finally, the institutionalization and review of lessons.These phases are more nuanced than the three phases typically referred to in the military innovation literature. 29Moreover, Pihs-Lang and Smith usefully distinguish between learning that may happen at theatre (tactical), headquarters (operational), and strategic (political) levels, either separately or in a synchronized and integrated way. 30The levels of learning, as well as the phases, are in reality not necessarily so distinct and sequential, but often overlap and blur-despite efforts of organizations such International Affairs 100: 1, 2024 as NATO to organize their learning in a stepwise process, as discussed by Dyson. 31et a more nuanced idea of learning in phases and their main focus is heuristically helpful for researchers to locate more precisely issues with learning performance.
In a second step, we distinguish between scope conditions, which we define as largely outside the control of organizations and their leadership, and organizational factors within their control.This distinction is generally not made in the CSDP learning literature and is also often subsumed in the military innovation literature under explanatory variables. 32It matters, because we are interested in correctly identifying and fairly evaluating learning performance and pathologies in our specific case of the battlegroups, not just to explain progress across phases.Making this distinction will be important for tracing learning over time in the battlegroup case.This allows us to highlight problems that were largely foreseen by expert communities at the time and those that were not and could not be, thus compensating against hindsight bias in post-mortems.In the identification of two of the scope conditions, we are influenced by Dunlop and Radaelli, who have differentiated between various modes of learning based on two variables. 33Which mode of learning prevails depends, first, on the degree of problem or issue (in)tractability-how technically difficult and uncertain an issue is from the perspective of decision-makers-and, second, on the certification of actors-the degree to which 'there is a sort of "teacher" that can be easily identified by learners and enjoys some social legitimacy'. 34e add two additional scope conditions.First, the strength and symmetry of signals of success or failure coming from the external environment of the organization.This aims to capture case-specific features that create strong incentives or pressures to learn.Depending on how symmetric these signals of success or failure are, they will influence how widely shared perceptions of failure or success are among the diverse decision-makers within an EU context.The strength and symmetry matter particularly in military CSDP, where decisions usually require unanimity and where member states differ in their strategic cultures, threat perceptions, domestic political contexts, and overall interests in building a strong CSDP.Conversely, weak and asymmetric signals from the environment will most likely hinder the emergence of shared perceptions.We argue that this condition is more appropriate to the EU context than the distinction between innovations during times of war and peace, which is frequently used in the military learning literature.The EU, as a whole, has never been at war, but one could expect strong signals to emanate from the actual or potential failure of a politically salient and resource-intensive CSDP operation.
The second additional scope condition is the prevailing political, economic and security context within which decision-makers and organizations operate.This is, again, largely out of their control.Particularly in the domain of military operations and missions, it may be influenced by prevailing threat perceptions in the aftermath of major attacks or a substantial improvement in relations to foreign countries or International Affairs 100: 1, 2024 regions.Political context could also be shaped by optimism or pessimism regarding the role of the EU as a military actor or the economic conditions that can enable or constrain member states' spending on security and defence.We agree with authors like Posen and Avant who emphasize the importance of political will and leadership as key factors for the implementation of innovations, but highlight that the EU's political context is more multifaceted compared to that of states. 35n determining the most important organizational factors that may help or hinder organizational learning, we are drawing on scholarly work on learning in the EU and NATO.As with the scope conditions, these factors could be either positive or negative and may fluctuate over time.For instance, Dyson stresses the arrival of new staff as a source of innovation and necessary disruption, whereas Hardt and Pihs-Lang argue that too much staff rotation and the prevalence of short-term contracts harm the creation of institutional memory and learning. 36In figure 1 we outline those organizational factors that promise the greatest explanatory power.
First, resourcing of learning processes captures the organizational priority given to learning capacities, while expertise focuses more narrowly on the institutional and thematic expertise needed through staffing.Second, processes and structures of learning capture the degree to which learning is codified and systematized and the processes and rules around learning across levels of hierarchy and between distinct units.Third, organizational culture captures the insight that learning is a profoundly social endeavour and is shaped by prevailing informal understandings, norms and (dis-)incentives that shape practitioners' understanding of which practices of knowledge production and diffusion are appropriate and beneficial to professional status and career progress.Finally, political leadership can instigate learning processes within an organization, help to push inconvenient and costly lessons through against resistance, signal encouragement, receptivity, disinterest or even hostility to the analysis of causes or suggestion of lessons by expert communities.
In the third step, we identify four potential learning pathologies as we move from a normatively neutral explanation of change through learning to a critique of potentially avoidable problems with organizational learning capacities. 37Resourcestarved learning is a pathology where organizations undermine their capacity to learn by not investing in specialized units that can move beyond routine organizational business in collecting and interpreting information related to organizational performance and its underlying causes, identifying actionable lessons and diffusing them.Low resourcing is an indication of low organizational priority for learning, and should be measured not just in terms of quantity, but also in terms of quality of staff.For instance, Hardt argues that a lack of training and awarenessraising on lesson identification and reporting has hindered NATO learning.38 International Affairs 100: 1, 2024

Figure 1: Learning phases, scope and organizational conditions and potential pathologies
Source: Authors' elaboration, building on literature review.
Disjointed and siloed learning is a pathology that arises when organizations do not manage to coordinate and integrate lesson-learning between functionally separate units or between tactical, operational and strategic levels.It leads to the absence of crucial information needed for understanding root causes or suffering from problematic attention biases, or even 'blind spots', in monitoring their environment.Bureaucratic politics within and between organizations creates a well-recognized impediment to such learning.Dyson rightly notes that scholarship on military learning stresses the need for 'well-organised learning processes' to 'ensure that learning does not remain "siloed"', for instance, through 'crossfunctional teams' comprised of all the services or consciously including civilians in these processes. 39epressed or curtailed learning is often the result of problematic organizational cultures, which might include a tendency to blame-shift or over-deference to hierarchy.This can implicitly penalize the reporting of errors and discourage necessary epistemic challenge to decision-makers and the communication of organizationally or politically inconvenient 'lessons'.For instance, Hardt found that NATO practitioners often choose not to engage with formal learning processes, either because of a fear of reputational damage when putting their name to observed lessons, or because of the 'need-to-know' information culture in military organizations.Political leadership can become a systematic problem for learning when it is uninterested and non-receptive to lessons identified.Conversely, it can also become a problem when political leaders stray beyond the legitimate bargaining and decision-making over which lessons to adopt by habitually interfering in the 'epistemic puzzling' phases to make sure lessons identified fit better with-or, at least, do not publicly jar with-salient political priorities.Political leadership can intersect with organizational culture when leaders gradually rise to the top, but it matters greatly in cases of politicians deciding on strategic-political lessons to be learned and internalized.

The role of learning in the evolution of EU battlegroups toward the RDC
In the following analysis, we first assess the EU's ability to learn across three distinct periods in the history of the battlegroups and the RDC: the initial conception of the battlegroups (2002-07); their standby period after reaching full operational capability (2007-16); and reform since the adoption of the EU Global strategy towards the creation of the EU RDC (2016-ongoing).We focus primarily on the actors, substance and appropriateness of the lessons identified and learned.We then investigate how the evidence fits the explanatory factors and pathologies discussed.

Artemis to EU battlegroups-2000-07
When the EU first expressed the ambition of creating a rapid-reaction force in 1999, the so-called Helsinki Headline Goal of developing a military corps-size capacity of 50,000-60,000 personnel was modelled on the size of the NATO-led implementation force (IFOR) in Bosnia and Herzegovina.Yet the EU only attained this goal on paper.The idea was further developed during the Franco-British summit held in February 2003 at Le Touquet, where a European rapid-response capacity was made a European priority.However, real progress was only made after positive operational experiences, which created a shared incentive for informal learning and an apolitical narrative of building on success among two lead nations.For the UK, the positive experience with the rapidly deployed military intervention (codenamed Operation Palliser) to Sierra Leone in 2000 played an important role as catalyst.It included a reinforced battalion group plus the UK Special Forces, supported by an amphibious ready group. 41For France, and the EU more widely, it was the rapid deployment of Operation Artemis to the Democratic Republic of the Congo in 2003 that particularly triggered progress.Operation Artemis consisted of 1,800 troops provided by 12 member states, with France acting as framework nation, to intervene for three months before handing over to a larger and longer-term UN organization mission, MONUC.Artemis offered the first real practical experience of what a rapidly deployed EU force could look like, and how the framework nation concept42 might contribute to that end.
International Affairs 100: 1, 2024 In the Franco-British declaration which followed the London summit of November 2003, the Artemis operation was welcomed as a success story.The declaration included a clear lesson learned: 'Together we now propose that the EU should aim to build on this precedent so that it is able to respond through ESDP [European Security and Defence Policy] to future similar requests from the United Nations …'. 43The declaration laid down the framework (battlegroup-sized forces, deployable within 15 days) of what would later become the EU Battlegroup Concept, proposed by France, Germany and the UK in February 2004, developed by the EU Military Staff tasked by the EU Military Committee, and approved by the European Council on 14 June 2004.Insights from a series of non-papers from 2004, made available to the authors, make reference to Operation Artemis as a blueprint for the EU Battlegroup Concept.
However, these non-papers also reveal discrepancies between political-strategic and military-operational learning processes.For instance, a national military official who has been involved in the development of the Battlegroup Concept from the beginning argues that Artemis only gave an indication of what the EU could potentially do in the future; in itself, the mission did not have significant or lasting positive impact.Another military official, adviser to his national EU ambassador at that time, emphasized how the Artemis experience was perceived at the political level: 'The Military Staff was asked to implement the political-diplomatic lesson, which was that Artemis was an operation the EU could and would do again in the future, and the battlegroups had to mirror this.' 44 He added: 'At the political-diplomatic level, they believed that every situation that required more than 1,500 troops and six months deployment would be addressed by someone else, by NATO.'The non-papers from 2004 suggest a similar politico-military discrepancy, especially regarding battlegroup size.In these non-papers, 1,500 troops was considered to be 'the generally accepted minimum force package' for missions-within the scope of the Petersberg Tasks, which guided the Helsinki Headline Goal of 1999-that have a rapid-response component and also need to include supporting elements together with strategic lift, sustainability and debarkation capability.In contrast, the Battlegroup Concept approved by the European Council regarded this size as a fixed goal or ceiling not to be exceeded, rather than a minimum floor on which to build.Likewise, the non-papers argued strongly that more than two battlegroups on standby for six months would be needed to mitigate the risk of potential national vetoes and competing crisis events.Yet, the approved Battlegroup Concept prescribes two on standby as the maximum.Therefore, the original design of the EU battlegroups was at least partially the result of learning from previous operations.However, the over-reliance on the Operation Artemis reference model and the discounting of advice on resourcing meant that the battlegroups were limited from the start in their utility to future crises with different or more demanding features.
capabilities.Other states plug their forces in. 43Franco-British Summit, declaration. 44 Another area for learning concerned the operational scenarios for which the battlegroups-and military EU crisis-management operations more generallycould be used.In the early 2000s, member states differed greatly in their interpretation of the operational scenarios prescribed by the Petersberg Tasks.Observers at the time noted that while such ambiguity 'may help to mask political differences, it is a problem for planning purposes, in particular for those working in the [EU Military Committee] and the [EU Military Staff ]'. 45The 2010 Headline Goal outlined a range of milestones to address this ambiguity and deal with capability shortfalls.The final text also stated that: 'Lessons learned from EU-led operations will also be taken into account.' 46Based on the experience of one of the authors of this article, 'illustrative scenarios' were at that time developed and agreed by the military at the EU Military Staff and EU Military Committee level for capabilitydevelopment purposes.However, the actionable lessons derived from this recognition of shortfalls did not remedy persistent ambiguity about operational scenarios at the political level.

EU battlegroups-2007-16
Many of the battlegroups' military operational shortcomings were recognized already from the point when they reached full operational capability by internal and outside sources in 2007.The new director-general of the EU Military Staff, General David Leakey, said that they needed: … a more agile command and control to manage them.What we have now is not bad, but could be improved.We also face the same problem as NATO with strategic lift: availability and funding of strategic lift are common issues for both organizations. 47ndstrom, an expert observer, recognized that '[g]iven its limited size and sustainability, an EU [battlegroup] is more likely to be deployed in the context of ongoing operations than operate independently'. 48He predicted that 'political pressure to employ an EU [battlegroup] is likely to increase with the passage of time', but highlighted that policy-makers are 'likely to look for very favourable conditions on the ground prior to the activation'. 49Lindstrom also noted pressure on national defence budgets as a reason for contributing countries 'to look for ways to avoid the activation of their EU [battlegroup] during a time of crisis'. 50urthermore, he critiqued narrowness of lesson-learning processes and recommended that '[policy-makers] should consider additional steps to encourage the International Affairs 100: 1, 2024 streamlining of lessons gathered across different departments and institutions to facilitate the formulation of more general sets of lessons learned'. 51n the years following the launch, it became increasingly difficult to fill the six-monthly standby roster.Continuous gaps opened up in the expected rota from 2012, as referenced above.These problems signalled decreasing levels of support for the battlegroups among member states and increased the risk of a battlegroup not being made available when called for.While some of these risks were anticipated at the conception stage, the greater puzzle is why it took the EU so long to publicly identify and implement lessons related to the problem's root causes.The primary reason why member states struggled to make the battlegroups work relates to political disincentives to use them.The prevailing political interests in cashing in on the 'peace dividend' grew stronger after the 2008/09 financial crisis.It strengthened the domestic role of finance ministries and fed into resource conflicts between service representatives in defence ministries, given that battlegroups were primarily land-based.During that period there was weak political leadership among member states for a strong EU role and a partially uninterested, partly distracted and overwhelmed new HR/VP, Catherine Ashton, overseeing the establishment of the European External Action Service (EEAS)-difficulties that are comprehensively covered in the literature. 52Widely discussed likely cases for battlegroup deployment-such as the post-election uprisings in Côte d'Ivoire in 2011 or the insurgencies in Mali (2013) and the Central African Republic (2013-14)-were not perceived to a sufficient extent as European (as opposed to French) problems. 53The consequences of non-deployment did not appear to be sufficiently large or immediate, and alternative ways of acting outside the EU structure were often found.Furthermore, there was little pressure from other EU leaders on battlegroup contributors to deliver on their commitments, because of generalized fears that such 'naming and shaming' would damage EU political coherence and support for future operations and missions.
On the funding issue, our research found overly narrow and partly contradictory interpretations of the root cause and how to tackle it.The Athena mechanism, created by a 2004 decision of the European Council on how to manage the financing of common costs of EU operations with military or defence implications, included provisions that guaranteed a periodic review-initially 'after every operation and at least every 18 months', later revised to every three years. 54owever, supervision and formal review of the Athena mechanism were placed under the aegis of Working Party of Foreign Relations Counsellors (RELEX) within the Council, which contributed to fragmentation of institutional responsibility for funding of EU peace and security measures-and risked contributing to disjointed learning.Pihs-Lang notes the institutional separation of the Athena 51 Lindstrom, Enter the EU battlegroups, p. 77. 52 reviews from other learning processes in the EEAS. 55Similarly, Nováky argues that Athena mechanism reviews repeatedly became bogged down in a 'diplomatic tug of war between France, the strongest supporter of expanded common funding, and the UK, its strongest opponent'. 56The 2008 financial and European debt crisis led to significant defence budget cuts across Europe, but 'different member states drew different lessons from Europe's economic problems', with some arguing for more common funding to improve burden-sharing, and others strongly opposing such plans. 57ore fundamentally, the 2014 Athena review showed how opponents to common funding referred to the lack of political will and strategic interests in conducting operations in Africa as the root cause of failure, rather than to frustrations about unequal financial burden-sharing.Athena mechanism evaluations centred narrowly on whether the mechanism worked according to the rules as described in the Council decision.Questions about whether these rules were sufficient to facilitate the use of a battlegroup were never formally part of these evaluations, because these were seen as political questions.The authors of this article are aware of instances when national military representatives from countries traditionally opposed to broadening the rules for common funding were actually sympathetic to such arguments.Yet they indicated that they would find it difficult to convince policy-makers at the political level-and indeed, any such advice was ignored when it came to Council discussions.For most of this period, the challenge of identifying the correct causes of member state reluctance was made more difficult by the provision-by member states themselves-of misleading explanations and unconvincing excuses, because the truth was seen as politically embarrassing in Brussels.For instance, Germany was opposed to shouldering a greater share of the cost based on gross national product, whereas the UK complained about having to pay twice, for its troops and for the common costs.Or, battlegroupproviding nations would come up with military-operational reasons for why their particular battlegroup was a poor fit for the crisis at hand, when in reality these reasons could have been addressed.

EU battlegroups to EU RDC-2016-23
European defence ambitions increased after Russia's 'annexation' of Crimea, the appointment of Federica Mogherini as HR/VP in 2014 and the outcome of the Brexit referendum in 2016.This combination of events led to increases in defence spending at the national level and a shift in EU strategy towards increasing the bloc's geopolitical power and military capabilities, first expressed in the 2016 EU strategy document known as the Global strategy 58 and, later, in 2022's Strategic International Affairs 100: 1, 2024 compass.These contextual and leadership changes gradually opened the door for more concerted efforts to identify some of the underlying problems and potential responses.For instance, in June 2017 EEAS secretary-general Helga Schmidt created a task force to review the EEAS's financial instruments, burden-sharing and harmonization, which resulted in a proposal from the HR/VP in June 2018 to create a wider European Peace Facility (EPF). 59As part of this new off-budget instrument, which replaced the Athena mechanism and the African Peace Facility, the financing of common costs of CSDP missions and operations increased from roughly 5-10 per cent to 10-15 per cent of the total costs.The COVID-19 crisis created a further impetus for strengthening organizational capacity for knowledge management and learning within the EEAS under the new HR/VP.A task force was created 'with members from the [Civilian Planning and Conduct Capability], [the EU Military Staff ] and "all other relevant actors" that started to collect lessons' from both the headquarters and theatre perspectives. 60Our interviewees suggested that the current HR/VP Josep Borrell (2019-) has been keen to promote integrative learning, including at political-strategic level, but doubts remain about the staff resources available for this purpose and whether a critical mass of member states is sufficiently engaged in this process.
The collapse of the Afghan national army and the chaotic US-led evacuation mission of 24-26 August 2021 prompted a search for lessons about what went wrong.It was also seen by HR/VP Borrell as a political opportunity to publicly make the case for creating a renewed rapid-reaction capacity-an idea we believe already existed at the higher military echelons in some member states: We need to draw lessons from this experience … as Europeans we have not been able to send 6,000 soldiers around the Kabul airport to secure the area.The US has been, we haven't.… For this reason in our Strategic Compass we are proposing the creation of a permanent European 'Initial Entry Force' that could act quickly in an emergency.… Our first entry force should be made of 5,000 soldiers that are able to mobilize at short notice.We have EU [battlegroups] but these have never been mobilized.We need to be able to act quickly. 61is 'Initial Entry Force' would later be referred to in the Strategic compass as the Rapid Deployment Capacity.Largely in parallel with the drafting of the Strategic compass, the EU Military Staff started working in 2021 on revisions of the battlegroups, based on lessons identified.The proposals were presented to the EU Chiefs of Defence meeting in May 2022, three months after Russia's invasion of Ukraine. 62The meeting acknowledged problems with the roster and agreed the following military-operational lessons to be learned: a default standby period of International Affairs 100: 1, 2024 12 months instead of six, staggered readiness, the Military Planning and Conduct Capability as the identified headquarters, more pre-coordination with framework nations, avoiding overlap in training demand, and advance planning which would allow for better scenario development. 63The Strategic compass stressed that increased modularity would make the RDC more flexibly deployable, made a commitment to providing strategic enablers, and outlined that initial development would be based on only two concrete operational scenarios based on reallife crises (rescue and evacuation; and initial phase of stabilization).This proposal would address the ambiguity problem discussed above and could help to make capacity shortfalls more visible, which in turn would increase the pressure to do something about it. 64oreover, the Strategic compass does not shy away from identifying and trying to address other types of root causes of failure, including a political commitment to extend the scope of common funding and to use more flexible decisionmaking arrangements, in particular the potential use of Article 44 of the Treaty on European Union.For instance, in April 2023 Johannes Hahn, the Commissioner for Budget and Administration, announced that the EU's first fully fledged live exercise will be financed on an ad hoc basis through the EPF.He also highlighted progress with regard to the use of Article 44 to allow coalitions of EU member states some more autonomy to plan or conduct an EU mission or operation. 65Yet, improvement of common funding through the EPF is not permanent.It is still dependent on a Council decision (by consensus) on a caseby-case basis.Resourcing the RDC may be deprioritized in favour of meeting the more politically salient NATO commitments, given the Russian threat.Furthermore, attempts at improving flexibility through Article 44 will neither remove the requirement for unanimity voting in the Council nor substantially reduce on its own the risk of potential national vetoes against the use of their military 'modules'.Although member states as a collective may want a more agile EU decision-making process for military missions, individually most of them do not want to give up the possibility of vetoing a mission or operation that may harm their interest.These interests continue to differ given variable threat perceptions, geographic interests and policy priorities that may not coincide with the crisis at hand.The problem of insufficient peer pressure on-or accountability of-those contributor nations that do not live up to their commitments made on paper is hardly identified, let alone addressed.
Furthermore, scarce resources still create obstacles to successful learning in the RDC development and its supporting architecture.One clear obstacle is the continuous pressure on the EPF budget caused by the military support to Ukraine.While the Council agreed on a significant EPF budget increase on 13 April 2023, concerns remain about how this financial pressure will affect the milestones set out in the Strategic compass (e.g. the further development and expansion of the International Affairs 100: 1, 2024 Military Planning and Conduct Capability, and the development of a Europeanlevel communication and information system).Interviewees at the higher EU military level noted how the current security situation-in particular, the war in Ukraine-overstretches the EU Military Staff, leading to reduced involvement in conceptual learning and development.Another largely unacknowledged problem so far is that financial burden-sharing is not just about increasing the share of common costs such as those for exercises.Member states worry mostly about the non-plannable costs, the additional costs of personnel on mission, transport, building of secure infrastructure, use of ammunition and fuels, significant higher maintenance costs, higher depreciation of equipment or even loss of equipment.This illustrates a wider problem of learning about unplannable future events that create incentives to let other countries shoulder the costs and risks.

Explaining and evaluating EU learning
Our discussion demonstrated that after years of delay, the EU eventually managed to identify and address some of the root causes of the creeping failure of the battlegroups through the RDC Concept.However, successful learning has related mostly to problems at the military-operational level.In contrast, it has struggled to fully diagnose or sufficiently address those causes that are more civilian, strategic or political in nature.Our theoretical framework helps to explain why some lessons are learned and others disappear completely or result in political ambiguity.Starting with the learning phases, we showed that the original conception of the battlegroups was informed by cases of perceived success which served as reference models and political arguments.Subsequently, for many years the experience of managing this new instrument created only weak or uneven perceptions of failure.The crises that triggered calls for the battlegroups' use generated only weak environmental signals of failure because they seemed too remote and small in their security or economic consequences to most member states.Furthermore, some of the potential negative impacts were avoided because individual EU member states acted outside the EU framework instead.The political salience of failure was further limited as member states largely refrained from publicly criticizing each other for blocking a mission.
Another hindering scope condition for learning was the intractability of the problem.The creation of highly prepared, effective and actually usable multinational rapidresponse forces is a novel challenge for which no successful 'off-the-shelf ' solutions exist.Few organizations have anything resembling such forces and the few that do are not fully comparable.For example, in the case of NATO this is because of the dominance of the US role and because the organization is unencumbered by imperatives of an 'integrated approach' involving civilian actors and instruments.NATO's track record for the use of rapid-reaction forces for crisis management cannot count as a success either.Moreover, designing a rapid-reaction force in a multinational setting is fraught with complexity, because success depends on understanding the interplay of diverse factors situated at different levels which International Affairs 100: 1, 2024 together create 'weakest links': The force needs to be militarily 'fit for purpose' across multiple dimensions, but will only be used if sufficient political incentives are in place.While certified expert knowledge exists to identify and address military operational problems, it is less clear who can authoritatively advise on the political disincentives.There is no easy way of avoiding parochial national interests getting in the way of mobilizing an instrument for common interests, as long as the key decision-makers are nationally appointed or elected and the Treaty on European Union requires unanimity in decision-making.
We also showed how a change in political, economic and security context matters to learning; first negatively, in the wake of the 2008 financial crisis, and then positively, after Russia's annexation of Ukrainian territory in 2014.Contextual change helps or hinders political leadership to construct functional imperatives and identify lessons to be learned.For instance, the Afghanistan evacuation operation of 2021 was objectively not a realistic candidate for an EU operation for a range of reasons, but it was politically salient and was framed by EU actors such as HR/ VP Borrell as a close call to underline the organization's lack of critical capacity and as a 'teachable moment' to mobilize support for change.Russia's full-scale invasion of Ukraine and the country's resistance with the unprecedented support of the US and EU since 2022 constitute another change of context with significant implications for learning conditions and political leadership.The invasion accelerated existing efforts to reform the battlegroups and improve the common funding of the RDC through the EPF.Conversely, it reduced funding for EU internal improvements, given the priority for support to Ukraine.Improvements in the EU will also suffer from the priority member states tend to give to investments to improve NATO's defensive capabilities.The war may furthermore distract political attention from the type of crises that the EU RDC is best placed to address in the EU's southern periphery.
We identified organizational factors which could be influenced by policymakers and senior officials.The political leadership of successive HR/VPs has mattered, both negatively and positively.While more recent political leadership through HR/VPs has improved the EU's capacity to learn, there is still an unfortunate tendency to sell renewed investment in EU instruments as efforts to 'learn' from specific politically visible operations.This can create an unhelpful straitjacket in terms of setting maximum troop numbers and constraining planning.It reverses the military-operational logic that required troop numbers should follow the needs of potential operations, not the other way round.It also increases the risk that a small set of forces designed for one specific crisis will have very limited applicability for future crises.The current HR/VP seems to understand the need to resource and create expertise for the lessons-learned process in the EEAS and the wider political environment, but is limited by restrictions on the creation of new positions because of budget limitations.
Differences in resourcing and expertise partly explain why the military part of the EU works reasonably well, because it is served by dedicated experts to run this process, including in the EU Military Staff.Although there is also a willing- ness to learn about security and defence on the civilian side, it is not formalized in a process, nor are there dedicated experts available to run through the learning phases.This is reflected and reinforced by differences in organizational culture between the military and the civilian sides, but also by informal norms shared across both.Military experts explained that in most military organizations a lessons-learned process is part of their standing operating procedures and part of military doctrine-notwithstanding some cultural differences between 'old' member states who have a longer tradition of transparent lessons-learned processes, compared to 'new' members affected by the legacy of the Warsaw Pact.In contrast, our interviewees noted career disincentives for civilian EU officials in the EEAS to identify and communicate inconvenient lessons.We found that the closer the process comes to the political level, the more difficult it becomes to discuss all topics to be improved.There is a strong consensus and 'face-saving' culture around military operations that hinders formally naming and shaming those partners that do not live up to commitments-much as Hardt found for NATO. 66National representatives, including the highest military representatives, generally wish to avoid their country being blamed for mistakes and seek to protect their national interests, including avoiding costs.In some cases, member states who vetoed or pushed back against formally recording certain lessons at the EU level have subsequently solved the issue nationally to avoid future criticism.However, this does not help to overcome the problem that sensitive lessons identified at the lower military levels often cannot be discussed and resolved at the highest international political levels.It explains, for example, why it took so long to get the issue of common funding on the political table in the EU, and how, even when this happened, it resulted in an ambiguous compromise.
The other obstacle to effective coordination of the military-operational and civilian-political levels of learning is situated in organizational structures.For instance, while the EU RDC Concept is officially a product of the EEAS and hence a responsibility of the HR/VP, the EU Military Staff Concepts and Capabilities Directorate holds the pen.This Directorate designs the RDC Concept and develops the modified EU battlegroups, but the HR/VP is responsible for coordinating with the member states and finding consensus.The Concepts and Capabilities Directorate can identify lessons at a higher strategic or political level, but these need to be dealt with by the appropriate higher authorities, including in the EU Military Committee, the Council and its subcommittees, and by the HR/VP.
Among the four potential learning pathologies, resources-starved learning is a problem, but arguably the least severe one.The EU rather remains prone to disjointed and siloed learning on problems that do not neatly fall into one sphere, despite improvement in recent years under the current HR/VP.Learning is indirectly repressed or curtailed by the prevailing organizational culture that prioritizes face-saving for the sake of maintaining political consensus and does not sufficiently reassure officials that they will not be blamed or punished for reporting about shortcomings.This can feed into uninterested learning among experts who International Affairs 100: 1, 2024 become frustrated by a lack of political interest or ambiguous political compromises.Still, the EU managed to improve its learning capacities in the present case and there is evidence of wider efforts, for instance, the creation of EEAS task forces on financing or COVID-19, and the consultative process leading to the publication of the EU Strategic compass.

Conclusion
The creation of usable and effective rapid-deployment forces is a difficult challenge for consensus-based regional organizations, especially for out-of-area crisis management operations.As shown in this article, they will only be used if the right military-operational, financial and political conditions are in place.We showed how the EU struggled for many years to correctly identify, let alone address, the root causes behind the creeping failure of its battlegroups.At the same time, we also argue that the EU has demonstrated its improved capacity to learn key military-operational lessons when designing the RDC: 1) The Battlegroup Concept was land-focused only, while the RDC will become a joint capacity; 2) Battlegroups lacked the support of earmarked strategic enablers with the same readiness, which the RDC will have; 3) The size of the RDC, although still limited, will be bigger than the battlegroups and fit for most of the foreseen tasks (except most initial entry operations); and 4) The EPF regulations foresee more common funding of missions.
Yet the EU's learning has been incomplete, as most of the deeper causes of the refusal of troop-contributing nations to meet their commitments remain either undiagnosed or unaddressed-for instance, in relation to the lack of reputational costs for reneging or remaining funding concerns.It is therefore uncertain whether the design changes will be sufficient for ensuring that the RDC will be used to good effect by the target date of 2025, especially given the resource competition created by Russia's full-scale invasion of Ukraine.
In line with recent writings by public administration scholars,67 our analysis demonstrates the merits of evaluating learning not just in terms of process, but also with regard to the substance and appropriateness of the lessons identified.Our findings align with the more recent military innovation and adaptation literature on the importance of organizational learning capacity, and particularly the role of culture.68Yet, our more nuanced conceptualization of learning phases and our distinction between scope conditions and organizational factors could be of value to this literature too.Our framework could also work for other multinational organizations, especially in a NATO context where the literature suggests similar : how the United Nations, Organization for Security and Co-operation in Europe and European Union have changed their crisis management institutions', Cooperation and Conflict 54: 4, 2019, pp.524-43, https://doi.org/10.1177/0010836718823814.Downloaded from https://academic.oup.com/ia/article/100/1/181/7506708 by Kings College Institute Of Psychiatry, user on 08 January 2024 capabilities19Pihs-Lang, Lesson (not) learned?, pp.14-15 and p. 154.20MichaelE. Smith, 'Learning in European Union peacebuilding: rhetoric and reality', Global Affairs 4: 2-3, 2018, pp.215-25, https://doi.org/10.1080/23340460.2018.1500427. 21Berlin Plus gave the EU access to NATO capabilities and assets for the Concordia and Althea operations only, but new EU member Cyprus's fraught relations with non-EU NATO member Turkey has served as a huge obstacle to cooperation between both organizations (cooperation which was already difficult) ever since.International Affairs 100: 1, 2024 40 39Dyson, 'The military as a learning organisation', pp.108, 121.40Hardt, 'How NATO remembers', p. 127; on organisational culture, see also Hoffman, Mars adapting, p. 126.International Affairs 100: 1, 2024 Interview with former national military official, 14 Feb.2023.Downloaded from https://academic.oup.com/ia/article/100/1/181/7506708 by Kings College Institute Of Psychiatry, user on 08 January 2024 International Affairs 100: 1, 2024 See e.g.Niklas Helwig and Carolin Rüger, 'In search of a role for the high representative: The legacy of Catherine Ashton', The International Spectator, 49: 4, 2014, pp.1-17.53Reykers,'Nosupplywithout demand'.54CouncilDecision 2004/197/CFSP establishing a mechanism to administer the financing of the common costs of EU operations having military or defence implications.