The EU’s truth by omission: Learning and accountability after the Eurozone crisis

Introduction

After major crises, national, international, and transnational organisations often take time to examine what went wrong. As the maxim states, a ‘crisis is too valuable an opportunity to waste’. Yet not all organisations seize the opportunity to learn, particularly those lessons related to their own institutional responses to the crisis. While the literature has developed theories explaining learning once it has occurred, the reasons for remaining resistant to certain lessons that could be drawn (potential learning) are unclear. For example, while a growing body of research has shed light on the economic and political lessons learned in the aftermath of the Eurozone crisis (Dinan et al., 2017; Schmidt, 2015; Zeitlin and Vanhercke, 2018), less attention has been paid to the relevant lessons that could have been included but simply fell off the radar.

To explore this interesting theoretical puzzle, we focus on the European Union’s (EU) decision after the Eurozone crisis to avoid setting up a truth commission as an independent institutional mechanism to convert past policy failures into lessons for future institutional reform. A caveat is in order: truth commissions are historically associated with societies emerging from authoritarianism and trauma and are intended to help them come to terms with the past and foster reconciliation and forgiveness. A good example is the South African truth commission (see Kovras, 2017). This article focuses on a different type of investigative commission, one designed to document institutional, political, and/or individual failures in the lead up to complex crises or policy failures and publish reports offering guidelines for institutional, policy, and regulatory reforms. ¹ As such, it can be an organisation’s key mechanism of learning after a crisis, and the EU might well have profited from its use. While scholarship has explored different levels of learning – individual/policymaker, policy, ideational (Bermeo, 1992; Chwieroth, 2010) – we focus on institutions of learning. We argue that understanding why the EU did not adopt a truth commission after the recent crisis reveals a lot about the nature of the EU itself and the scope of the lessons learned, including why some but not others were learned.

Setting up an institutional mechanism of learning has helped a number of countries come to terms with a crisis, strengthening their institutional responses to future crises. In the United States, for example, the ‘Pecora Commission’ not only identified the causes of the 1929 Crash leading to the Great Depression but also recommended institutional reforms that resulted in the protection of the markets from another financial crisis for decades. More recently, days after the collapse of 97% of the Icelandic banking system in 2008, a very successful independent ‘Special Investigation Commission’ (SIC) was set up to document the causes and recommend reforms to Icelandic state institutions (Hjalmarsson, 2017). After a detailed investigation, in late 2010, the SIC released a 2200-page report revealing systemic flaws in state institutions and the banking sector – it became an instant best-seller. Parliament adopted its recommendations and converted the reform proposals into legislation ranging from banking regulation to ministerial code of contact and institutional culture (Government of Iceland, 2015; Hjalmarsson, 2016). By offering an evidence-based/authoritative narrative of the causes of the financial crash, the commission delimited the number of permissible lies circulated by demagogues and spoilers who could have revised and/or denied fundamental aspects of the crisis. ² Similar commissions of inquiry, albeit less successful, were set up in Ireland and Cyprus (Kelly, 2017; Pegasiou, 2017). As Boin et al. (2017: 18) note, ‘Rendering account of what happened, and why, is a crucial task’ in a democratic institution. Another consideration is that from the perspective of accountability, the technical aspects of contemporary economic crises are frequently so complex that they are fully understood only by a small number of experts. Therefore, truth commissions have gained currency as mechanisms establishing a simplified, yet authoritative, narrative of the causes of a crisis that can be easily understood by the broader public (Boin et al., 2017).

The absence of such an independent commission to convert past policy failures into institutional lessons in instances when it would clearly help poses a puzzle to the organisational theory of learning. Organisational theory posits intelligent organisations have the capacity and the interest to learn from failures to improve their legitimacy and their performance. From this vantage point, a crisis, such as the Eurozone crisis, represents an opportunity for organisations, such as the EU – especially if it forces the organisation to investigate the (mis)management of the crisis and recommend improved responses in the future. Another international organisation vested in the Eurozone crisis, the International Monetary Fund (IMF), launched an independent investigation through its Internal Evaluation Office (IEO). The investigation identified flaws in the Fund’s response to the crisis and offered recommendations to improve its engagement in future crises (IMF, 2016). A similar independent investigation did not take place at the EU level.

Oddly enough, the EU had set up fact-finding bodies in the aftermath of previous policy failures. For instance, it established an independent inquiry board to investigate causes of the ‘Galileo Incident’, Europe’s satellite project, and the board generated policy recommendations to avoid new failures (European Commission, 2019). More recently, at the international level, the EU led a proposal for an independent investigation of the origins of the COVID-19 pandemic (Scott et al., 2020).

In what follows, we clarify our understanding of learning and assess three broad theoretical approaches to its study – rationalist, ideational, and organisational/institutional explanations – in an effort to find the most convincing answer to the main question of this article. After describing our research design, highlighting the inherent obstacles to investigating counterfactuals, and ruling out alternative explanations, we offer our main argument. We argue that the organisational design of the EU, notably the absence of a permanent ‘meta-learning’ body, has locked EU institutions and policymakers into a ‘reactive mode’ of learning. In effect, during a heavily politicised crisis, all such organisations and policymakers switch to a ‘fire-fighting’ mode where the scope of relevant lessons is narrowly defined to suit the effective response to the crisis. Not all organisations are similarly constrained. Organisations with embedded but independent institutions of meta-learning – bodies that monitor their performance away from the crisis – follow a ‘proactive mode’ of learning, making it more likely for them to learn broader lessons from policy failures. The organisational structure decouples the independent/technical investigation of failures from the heavily contested public debates around issues of responsibility and accountability. In the absence of this institutional capacity of the EU to ‘learn how to learn’, the scope of relevant lessons has remained partial, confined to the management of the economic crisis (i.e. economic policies or certain path-dependent institutional reforms), precluding good practices in other areas like accountability and transparency.

A final conceptual caveat is in order. ‘Accountability’ is the uber concept of the early 21st century, to use Matthew Flinders’ (2014) apt remark, and the ‘golden concept’ according to Bovens et al. (2008). Despite its growing prominence, it means different things to different scholars, ranging from the ‘vertical’ accountability required by office holders who want to win re-election, to ‘horizontal’ mechanisms – such as audits and inspections – geared towards promoting transparency in public office (Bovens, 2010; Crum and Curtin, 2015; Hilliard et al., 2020; Papadopoulos, 2010). Our approach to the study of accountability entails two core elements that overlap with the mandate of truth commissions: retrospective and prospective accountability. Retrospective accountability includes accounts of past actions or omissions in the lead up to a crisis and its (mis)management. Prospective accountability includes the introduction of reinforced regulatory frameworks to increase the transparency and legitimacy of the organisation’s institutional response to future crisis (Schmidt, 2020).

Theories and concepts of learning

Learning is an elusive concept and theories of learning abound. Rationalist theories maintain that in normal times, any change to the status quo incurs costs, and the costly endeavour of learning makes actors and organisations reluctant to engage in self-reflection (Rose, 1991). But in times of crisis, opportunities for learning may outweigh the costs (Haas and Haas, 1995; Hall, 1993; May, 1992). Ideational theories argue exogenous shocks catalyse learning, but individual policymakers intuitively seek information and facts from the real world that will reiterate the ‘prevailing’ ideas of a particular organisation (Chwieroth, 2010). In effect, what policymakers identify as ‘relevant’ and ‘useful’ shapes the direction of learning after a crisis (Chwieroth, 2010). Scholarship from a wide range of sub-domains in political science, from international political economy to democratisation, to name only a few, concurs that crises serve as a learning catalyst at the individual, policy, and ideational levels (Bermeo, 1992; Blyth, 2001).

Although it is now conventional wisdom that crises stimulate political learning, limited attention has been paid to the institutional mechanisms steering this process. Arguably, ‘different types of learning occur during crises, […] shaped, at least partially, by the institutional mechanisms established to drive the process’ (Kovras et al, 2018: 177). If so, examining the institutional mechanisms established to tackle a crisis can indicate whether and how international/transnational organizations learn from the past, and the scope of (relevant) lessons learned. The relevant literature has not sufficiently explored this possibility. To begin to fill the gap, we focus on one particular institutional mechanism (truth commissions) specifically designed to convert failures into lessons. Although it is true that different levels/types of learning may overlap, we focus on institutions, ‘believing the institutional mechanics of learning can reveal a great deal’ about other types of learning ‘by illuminating the political drivers shaping policy responses’ (Kovras et al, 2018: 177).

Theories of organisational learning are helpful to this type of exploration. While rationalist and ideational theories tend to focus on the level of individuals (policymakers), ideas, and policies (programmes), the organisational literature emphasises the institutional mechanisms and routines shaping the prospects of and the obstacles to organisational learning. Instead of focusing exclusively on the outcome of learning, this literature emphasises the mechanisms and routines of learning (Scott, 1995), arguing they provide ‘actors with a given set of cognitive categories and a typology of action options’ (Tsoukas and Vladimirou, 2001: 978). In this view, the institutional mechanisms and norms within an organisation shape individual knowledge and learning. Yet organisational learning is more than the sum of individual learners; policymakers may learn at the individual level, but this alone is insufficient to trigger learning at the organisational level. The causal chain goes in the opposite direction, highlighting the importance of organisational mechanisms and norms and their shaping of the ideas, actions, and knowledge of individuals. Research in the area of behavioural studies indicates actions in an organisation are shaped by the logic of appropriateness and a need for legitimacy, not simply by the logic of consequences or reactive learning (Levitt and March, 1988: 32; Nelson and Winter, 1982). Finally, as noted above, an intelligent organisation has the capacity to be self-reflective and learn from its own failures to improve its performance and legitimacy.

This points to a different type of learning uncharted by other schools of thought. The literature makes a useful distinction between ‘single-loop learning’, referring to adjustments to respond to pressing external problems, and ‘deutero-learning’, referring to a more fundamental process of rethinking and revisiting fundamental beliefs (theory of action) of the organisation (Argyris and Schon, 1974: 30). These categories generally correspond to what ideational and rationalist theories label ‘adaptation’ and ‘learning’, respectively. Yet organisational theories highlight a third level of ‘meta-learning’, namely the institutional mechanisms organisations set up to ‘learn how to learn’, both in normal times and in times of crisis. We see truth commissions as the closest approximation of the institutional mechanism of ‘meta-learning’, or the mechanism of converting past policy failures into future lessons.

The reaction of the EU after the Eurozone crisis is puzzling to all schools of thought. First, at the most fundamental level, an organisation tasked to deal with ‘existential’ crisis is forced to investigate the causes of the crisis and come up with specific recommendations. Hence, a commission of inquiry – or a similar backward-looking mechanism – to identify the sources of the problem could be expected to take place, but this did not occur. Instead, the response to the crisis remained limited to ‘single-loop’ learning, that is, tweaks to macroeconomic policy and some path-dependent institutional reforms, mainly prioritising the preferences of European corporate actors (see below). Second, it queries organisational theory: why did the crisis not result in the establishment of an inquiry commission that would review the wrongdoings of the past and offer policy recommendations? Perhaps an independent evaluation office could have been created. Apart from being good practice, it would have boosted the external legitimacy and accountability of the EU institutions in the eyes of European citizens – major issues after the Eurozone crisis. In what follows, we attempt to solve the riddle.

Research design

Raising and addressing counterfactuals, as we do here, is a challenging – almost impossible – task, but if executed properly, it can benefit theory testing and refinement (Cappoccia and Kelemen, 2007). To reduce the risks associated with the study of counterfactuals, we focused on the ‘most likely case’ (Eckstein, 1975), in this instance the organisation most likely to offer lessons related to accountability (Hilliard et al., 2020), for reasons explained above. In addition, we systematically reviewed all sources where we would expect to find the key lessons learned by the different bodies of the EU during the crisis, thereby mapping the framing of the causes of the crisis. These sources included the Economic Adjustment Programmes (EAPs) designed by the Troika (IMF-European Central Bank-European Commission) as part of the EU’s financial assistance to five crisis-ridden member states (Greece, Spain, Portugal, Cyprus, Ireland). We also analysed the European Commission’s regular review reports on the bailout countries, as these assess whether the terms of the adjustment programme are met. ³ Similarly, we scrutinised the reports of the European Commission and European Parliament to determine how EU institutions framed the causes of the crisis and developed policies to protect member states from future crises. Finally, we documented major policy and institutional reforms implemented by the EU post-2010 to explore the types of learning at the EU level. These steps allowed us to map whether issues of accountability were featured in the intra-organisational debates and policies after the crisis. Yet proving an issue was not on the table cannot overcome an overlapping methodological problem, namely, how to study ‘silences’ in particular policy domains. To this end, we took two further steps.

First, we conducted 11 semi-structured in-depth elite interviews. Following Tansey (2007: 766), we relied on a ‘non-probability sampling’ to establish what a set of people who directly participated in post-crisis reform processes in the EU thought about independent commissions. We interviewed very senior people representing different EU institutions and political elite of the EU member states. Interviewees included members of the European Parliament involved in the economic governance reforms, senior officials in the European Commission who took part in the Troika missions and contributed to the design of the reform packages, and a very high-level official representing Eurogroup. We also interviewed lawmakers and officials across the crisis-ridden member states (see Appendix 1). We interviewed each person once; in some cases, we asked follow-up questions by email. Interviews lasted around 20–30 minutes. We expected interviewing high-profile policymakers with direct exposure to the experience of countries using the policy instruments of interest would make it easier for us to determine whether the EU discussed these as useful or relevant. Second, in addition to triangulating interview data, we tested alternative explanations from the literature. When these failed to come up with a convincing account, we developed our own argument, drawing on theories of organisational learning. Throughout the process, however, we remained mindful that addressing ‘the dogs that didn’t bark’ (the lessons not learned) is inherently challenging.

Argument: Absence of self-reflective organisational capacity and EU’s partial learning

The Eurozone crisis was an instructive period for EU institutions in particular and the EU integration project more broadly. It is true that the EU learned considerably from the Eurozone crisis and implemented a certain set of reforms, such as strengthening Stability and Growth Pact, developing macroeconomic surveillance framework, and completing the Banking Union (see Copelovitch et al., 2016; Juncker, 2015; Verdun, 2015). The European Commission and other EU institutions also commissioned discussion papers from independent experts on the causes of the euro crisis and issues of post-crisis legitimacy and accountability (cf. Alcidi et al., 2014; European Central Bank, 2018; European and Commission, 2009a; Schmidt, 2015). Still, as we explain below, the lessons were ‘partial’, focusing on economic policymaking in an effort to deal with the crisis, sidelining other important lessons about the operation of the EU itself.

The Eurozone turmoil, which broke out in 2010, threw the entire European integration project into an existential crisis, and the economies of several member states came to the brink of collapse (Dinan et al., 2017). In response, ‘policy’ and ‘institutional’ learning took place in the EU to sustain single currency. On the policy side, wide-ranging reforms were introduced to improve fiscal discipline and accountability (Copeland and Daly, 2015; Juncker, 2015: 22), by scrutinising national budgets and sanctioning those failing to comply with the fiscal targets (Crespy and Menz, 2015; Verdun, 2015: 228; Warren, 2018; Zeitlin and Vanhercke, 2018: 150). The EU also introduced a set of institutional reforms to mitigate imminent accountability problems emanating from the flawed institutional architecture of the Economic and Monetary Union (De Grauwe, 2011; Hall, 2012; Jones et al., 2016). In effect, the incomplete nature of the euro area, which ECB president Mario Draghi (2014) called the ‘Achilles heel’ of the European integration project, exacerbated the accountability problems. Therefore, the EU focused on regulatory reforms to obtain oversight of the banking system and better regulate credit rating agencies, European securities market, and the pension system (Juncker, 2015: 11).

Given that the ‘idea of crisis as opportunity [is] so deeply woven’ into the history of European integration (Dinan, 2017: 16), the policy learning and lesson drawing post-euro crisis is not all that surprising. However, it is strange that the EU learned important fiscal and financial lessons but did not widen the learning focus. An independent inquiry commission could have turned a spotlight on areas of the Eurozone crisis that were underexplored in the context of reactive learning. First, it could have identified the broader causes of the Eurozone crisis, namely the institutional, economic, and other factors that left the EU, not merely its member states, vulnerable to external shocks. Providing a master narrative of what went wrong based on a systematic investigation of the past is an important form of accountability and, as such, would consolidate EU legitimacy. Second, it could have offered a systematic evaluation of the policy responses to the crisis and their overall impact. In what areas were the programmes co-designed by the Troika effective, and where did they fail? Evaluating the policy responses co-designed by external/independent organisations (such as IMF and ECB) inevitably entails some dilution of responsibility, inhibiting a proper investigation and attribution of blame. It is precisely for this reason that such commissions would be beneficial for the EU. They would provide a reflexive platform from which to raise and address the big questions. Was the cooperation with other organisations or bodies useful? Would it be advisable in the future and, if so, in what areas? What could the EU do differently in a similar crisis in the future? As opposed to ad hoc learning in specific domains, these questions necessitate systematic reflection on the organisation itself. It is hardly surprising that despite facing the dilemma of convoluted responsibility, neither the EU nor individual countries set up commissions to learn lessons for the future. Yet such commissions could have helped generate concrete policy recommendations about future crisis management.

Interestingly, it was the EU’s monitoring of bailout programmes that developed the instruments used by some of its member states to look for answers. Moreover, leading EU policymakers publicly acknowledged the benefits of inquiry commissions; some even recommended their adoption as a good practice. For example, Thomas Wieser, President of the Eurogroup Working Group (EWG), highlighted the potential benefits of an investigative truth commission in countries facing economic challenges (Athanasiou, 2018). Other policymakers had experience directing similar fact-finding mechanisms in individual countries. For instance, Klaus Regling, managing director of the European Stability Mechanism, led a committee of inquiry in Ireland early after the crisis (Regling and Watson, 2010). In short, this mechanism was understood to be both relevant and valuable, but this knowledge did not result in lessons travelling from the national to the transnational level.

There is another reason why we might have expected a probe at the EU level. As the literature on crisis management shows, after a major disaster, the organisation involved usually creates backward-looking mechanisms to draw lessons and improve its performance to avoid repeating the same mistakes. After a plane crash, aviation authorities launch an investigation to identify the causes, learn from the flaws in the system, and come up with more effective procedures to prevent similar disasters in the future. The Eurozone crisis had enormous ramifications. And there is no guarantee that a similar predicament will not happen again. In this context, the EU’s failure to set up an independent investigation or even consider adding a discussion of its utility to its agenda stands out. As mentioned above, the EU’s emergence and expansion are inextricably linked to crises, and the idea of a ‘crisis as an opportunity to learn’ is a prominent feature of EU institutions. Yet no official inquiry commission was tasked with documenting the causes of the crisis.

We advance an organisational argument to begin to unravel the mystery. Drawing on the pioneering work of Argyris and Schon (1948), we argue that in the aftermath of a crisis, there are two pathways to learning: reactive and proactive (see Graph 1).

OPEN IN VIEWER

Graph 1.

Two varieties of EU learning and accountability.

The reactive learning cycle suggests that in times of crisis, policymakers and organisations are trapped in a fire-fighting mode. Their main objective is to develop policies to recover, and their learning capacity is mostly limited to issues overlapping with the obvious causes of the crisis or, in other words, to simply to put out the fire. In the aftermath of a complex politicised crisis, most organisations are trapped in this reactive learning cycle. Debates about the causes and the (mis)management of the situation are often derailed, and a political blame game to score political points begins, thereby reducing the scope of lessons that can be learned to the level of the individual policymaker. The organisation itself engages in a process of blame deflection, particularly in heavily politicised crises (Hood, 2011), making it easier to launch investigations into technical issues that do not raise significant questions about political responsibility.

This reactive ‘single-loop learning’ epitomises the EU’s approach to learning after the recent crisis. The EU confined itself to selective policy adjustments and targeted institutional reforms in the economic domain (banking union) and excluded the possibility of learning broader lessons on democratic legitimacy, restoring trust in EU institutions, or acknowledging accountability. In this type of learning, policymakers and organisations are trapped in a struggle for survival, leaving little room for self-reflection, even during a crisis (Boin et al., 2017). But this should not be equated with a deterministic account of a crisis as inevitable. Nor do all organisations react the same way in a crisis. And not all crises yield the same learning outcomes.

We argue that only organisations with an embedded institutional capacity for self-reflection (‘meta-learning’) possess the required institutional means to identify good practices not directly related to the crisis and convert past policy failures into institutional lessons for the future. This proactive learning cycle necessitates the presence of an independent agency in the organisation that will systematically review the past performance of the organisation in an effort to identify both good practices and areas needing improvement. This type of meta-learning institution has the potential to put the spotlight on issues that would otherwise remain off the agenda of organisations trapped in a fire-fighting mode. Most importantly, only organisations with this meta-learning capacity can overcome the politicisation trap by decoupling the independent/technical investigation of policy failures from the heavily contested public debates of responsibility and accountability. In short, it provides institutional insulation from the post-crisis blame game.

Based on this logic, in the absence of a self-reflexive organisational capacity (or a mechanism of ‘learning how to learn’), we argue the scope of useful or relevant lessons in the EU remained limited to crisis management, precluding good practices in other areas. To be precise, we do not suggest such an independent office would have inevitably catalysed such an outcome. We only highlight that, in its absence, it becomes almost impossible to expect policymakers to disengage from reactive learning.

To probe this further, we need to take a closer look at the EU’s organisational structure. We maintain that the EU’s institutional structure made it less likely to draw lessons from either its own failures or the good practices of its member states. Three major mechanisms of its atypical organisational structure impeded post-crisis learning.

First, as Van Esch and Swinkels (2015: 1204) underline, the EU is characterised by a ‘strongly fragmented governance structure’. There is no well-identified institutional authority to take the initiative at critical junctures and serve as a critical ‘focal point’ to collect information, assess causes of the problems, frame a coherent crisis narrative, and develop reform proposals (Kamkhaji and Radaelli, 2017: 717). The crisis caught European ruling elites by surprise. Massive uncertainty about the causes and confusion about the steps to be taken dominated the policy agenda as immediate problems. The European Commission, the Council of the EU, the ECB, and member states adopted separate – and often contradictory – narratives about the unfolding calamity (Henning, 2017). In a fragmented governance framework, the absence of meta-learning institutions, such as an independent office to evaluate the performance of the EU’s programmes, led to the emergence of multiple crisis narratives and impeded coherent response. Instead, EU officials developed two types of discourse as an outcome of the blame- and responsibility-shifting at the domestic-EU nexus: first, accountability is traditionally the prerogative of the national sovereignty of individual member states and does not fall into the remit of activities of the EU acquis communautaire; second, tools of accountability can be used to score political points, increasing polarisation and politicisation. In short, policies of accountability are seen as irrelevant and could unnecessarily divide the EU.

We heard both arguments from the policymakers we interviewed. One interviewee said the European Commission focused ‘more on changing the economic framework’ in the EU, not accountability issues and inquiry commissions. ⁴ Identifying those responsible for the crisis was considered an out-of-mandate issue to be dealt with at the national level. The EU institutions hesitated to get involved because they didn’t want ‘to breach national sovereignty’. ⁵ Concerns about mandate and sovereignty led them to see national parliaments and national courts as the proper authorities to investigate those responsible for the crisis and prosecute individuals. ⁶ In effect, threat and urgency shaped the ad hoc policy responses of the EU institutions and policymakers (Van Esch and Swinkels, 2015), as a result of which, do-or-die reforms crowded out serious concerns about post-crisis accountability. ⁷

Second, the design of the reform agenda further complicated the already ‘tangled governance’ structure of the EU (Henning, 2017). In the wake of the euro crisis, the Troika became the main institution designing and supervising reform programmes in crisis-ridden countries. However, any assessment of the management of the crisis would require the evaluation of the functioning of several different supranational actors, including the ECB, the European Commission, or even the ESM at later stages in the crisis. As Henning (2017: 9) notes, these actors have ‘distinct memberships, missions, and capabilities and they negotiated among themselves as they bargained with the governments that borrowed from them’. Each has different understandings and mechanisms of accountability, and none has the authority to investigate others. The only actor able to set up a commission is the Council of the EU. Given the northern (creditor) and southern (debtor) countries’ radically different agendas, interests, and understandings of the causes of the crisis, it cannot be a surprise that this was never even discussed. In a nutshell, the institutional structure and the lack of organisational capacity for proactive learning precluded an EU commission of inquiry.

The multifaceted institutional responses and blame games at the domestic-EU nexus complicated governance problems by fragmenting the political agenda and reinforcing a reactive policy learning cycle that pushed for incremental path-dependent changes, not a paradigm shift. This emerged in interviews with senior national policymakers with exposure to EU operations during the crisis. For example, a former Greek Minister of Finance acknowledged that a systematic backward-looking investigation ‘would be very useful’, as it could draw a line between the ‘IMF that was the only institution that had such a mechanism of self-reflection . . .’ and the ‘European Commission, the ECB and most importantly the Eurogroup [but this] never happened’. ⁸ In explaining this outcome, he highlighted the perception among EU policymakers that ‘issues of accountability rely almost exclusively on state sovereignty’ (ibid).

Third, the EU governance architecture feeds into a complacent view that the EU already has multiple accountability mechanisms. Our data suggest EU policymakers are inclined to believe the EU has better accountability mechanisms than other comparable international institutions. Accordingly, the EU is considered to have several internal retrospection and external control mechanisms. A senior European Commission staff member named the European Court of Auditors (ECA) as the external control authority. ⁹ However, ECA cannot be compared with an independent evaluation office, as its mandate is limited to ‘improve the European Commission’s management of the EU budget and reports on EU finances’. ¹⁰ During the Eurozone crisis, this complacency oriented the debate towards improving the accountability and transparency of already existing institutions, rather than finding new self-reflective institutions that could have shaped the discourse, formulated new ideas, and promoted out-of-the-box thinking. ¹¹ Instead, debates centred on the democratic oversight of the European Central Bank, the inclusion of the European Parliament in economic decision-making procedures, and the consequences of the Troika programme (Crum, 2018; Dawson, 2015; Verdun, 2015). Ironically, the way the EU framed post-crisis accountability narrowed the space for alternative ideas and further reduced its already limited self-reflective capacity.

Contravening the international experience

Our findings have significant policy implications. The embedded capacity of organisations to create mechanisms that can convert failures to lessons is considered good practice both in practical policymaking and in theory (Gutner and Thompson, 2010). Most international organisations have embraced this practice. For example, the World Bank set up the Operations Evaluation Department (OED) to evaluate the performance of programmes designed by the Bank and its affiliated agencies, in an effort to find flaws and designate areas for improvement (learning) (Grasso et al., 2003; McKay, 2003 Woods, 2001). The IMF followed suit with the establishment of the Independent Evaluation Office (IEO) in 2001 (Lissakers et al., 2006; Weaver, 2010). With the Asian financial crisis in the 1990s, the IMF had come under significant pressure from donor countries to create the IEO to boost its external legitimacy and effectiveness (Avant et al., 2010; Seabrooke, 2007: 258; Weaver, 2010: 367).

Alternative types of learning

We argue that although rationalist and ideational approaches offer valuable and complementary insights into the EU’s selective learning, organisational learning provides a comprehensive account of its failure to develop a coherent policy on accountability. Rationalist approaches predict very minimal learning, and this learning takes an incremental and reactive form at best, as policymakers are inclined to minimise costs of reforms. These arguments were popular among the EU policymakers we talked to, yet they only partially explain the situation. For one thing, the EU demonstrated significantly more reform activism than the ‘minimal threshold’ expected by rationalist approaches – especially in later stages of the euro crisis. The Troika acted assertively when intervening in domestic affairs, including on politically sensitive issues, particularly in crisis-ridden bailout countries, so much so that serious concerns were raised about the erosion of democratic legitimacy in these countries.

Signing a bailout deal de facto is an encroachment on national sovereignty, as the bailed-out state agrees to relinquish its autonomy in fiscal policy decisions. The question is more about the degree of EU meddling with national affairs than the principle. In-depth analysis of EAPs suggests the Troika took an active stance on delicate political issues. In the Greek case, for instance, the Troika deployed a proactive position on issues traditionally falling into the remit of the national judiciary and legislature. Probably, the starkest illustration is its open support for the former chief of the Greek statistical authority who was under investigation by the judiciary; in fact, monitoring developments on this case became part of the official conditionalities and was subject to ‘enhanced surveillance’. ¹² The Troika also intervened in the following domains: ‘reporting corruption in tax administration’ (European Commission, 2010: 100), the ‘development of full-fledged anti-corruption plan’ (European Commission, 2012: 37, 79–80), the formation of a ‘corps for the prosecution of financial crimes’ (European Commission, 2013a: 12), and the ‘appointment of a national anti-corruption coordinator’ (European Commission, 2013b: 47). These and other interventions consolidated the ‘troika’s super sovereign status’ (Watkins, 2012: 12).

Thus, even though accountability was not part of the hard conditionality of bailouts, arguably because of concerns about national sovereignty, the Troika openly intervened to shape the direction of reforms on issues that traditionally fall into the realm of national sovereignty. More importantly, even if concerns about national sovereignty could explain why the EU might think twice before adding accountability to the agenda, they do not explain why fact-finding was not endorsed as part of its non-binding good practices for countries facing similar challenges in the future.

Ultimately, although rationalist accounts explain ‘selective’ and incremental post-crisis learning, they are less able to explain why the EU did not set up a fact-finding commission to learn from its own (organisational) flaws. Most organisations confronting failure or crisis – from environmental or aeronautical disasters to political fiascos – launch independent investigations (Boin et al., 2017). Despite facing one of the most severe crises since its inception, the EU did not do so, challenging rationalist perspectives. Pressures for accountability came not only from debt-ridden countries, with protests in Athens or Madrid, but also from creditor countries like Germany, where taxpayers quite rightly wanted to know why they had to pay for bailouts in other member states. Surely the cost of establishing an independent truth commission is minimal compared to the potential political costs of refusing to address accountability, including growing distrust in EU institutions.

The literature confirms that politicisation conditions post-crisis policy learning (Birkland, 2006; Dekker and Hansén, 2004). As Broekema (2016) demonstrates, the ‘politicization of issues either promotes or impedes crisis-induced EU learning’, depending on the intervening factors. From this vantage point, we might hypothesise that the politicised nature of inquiry/truth commissions in the EU prevented this mechanism from appearing on the EU agenda after the Eurozone crisis. Our interview data also indicate that the European Commission hesitated to take a position because truth commissions in some member states evolved into political witch hunts and blame games. ¹³ However, this argument only partially explains the EU’s limited learning because of the selection bias it entails.

First, for every commission that was instrumentally deployed to score political points, there was a successful one that documented the causes of the crisis and offered useful recommendations for institutional reform (Kovras et al., 2018). If we subscribe to this account, we would expect at least some intra-organisational debates on the political and other costs of embracing lessons related to accountability. Yet our perusal of the EU publications, coupled with our interviews of high-profile policymakers, paints a different picture. There was a complete silence around the prospect of deploying a fact-finding commission of inquiry, and the issue was never discussed in Troika meetings. ¹⁴ This was not a rejection after meticulous internal evaluation – it was simply never on the table.

Second, an established literature on post-crisis accountability suggests the contentious and politicised nature of a crisis is the ‘arena in which politicians and stakeholders struggle over causes and blame’ (Boin et al., 2017: 105). According to this literature, it is precisely the politicised nature of the crisis that enables questions about responsibility and blame (Boin et al., 2017; Kuipers and ‘t Hart, 2014), creating opportunities for accountability mechanisms to be employed. In other words, retrospective policies of accountability offer an opportunity for powerholders to deflect blame (Hood, 2011) and retain the support of unhappy citizens by conveying principled politics and their endorsement of the rule of law.

We still need to explain why accountability was completely side-lined at the European level. This brings us to the role of the ideas that dominated the post-crisis learning repertoire. Ideational explanations are also well suited to explain why certain ideas (like fact-finding commissions) were seen as irrelevant and excluded from the official agenda of the EU. This line of reasoning suggests the ‘prevailing ideas’ in an organisation shape the modes of learning and policy change. However, as discussed already, a crisis can be a game changer that delegitimises the prevailing ideas and provides opportunities to introduce new ones to create paradigm change (Bermeo, 1992; Blyth, 2001; Chwieroth, 2010; Hall, 1993).

Arguably more challenging for ideational accounts, as we noted earlier, individual high-level EU policymakers publicly endorsed inquiry commissions as relevant and as good practice. Thomas Wieser, President of the Eurogroup Working Group (EWG), stressed:

In Ireland a commission of inquiry to investigate the causes in the lead up to the crisis, they reached certain conclusions and some individuals ended up in jail. The same happened in Spain and in Cyprus and today we know what led to the events of 2013.

He concluded:

‘The only countries that nothing has happened are Greece and Italy. In Greece there is no official apportion of responsibilities; therefore, it is difficult for the people to understand what happened and to plan for the future or to protect the country from repeating the same mistakes (Athanasiou, 2018).

Also mentioned previously is that other EU policymakers had direct experience leading fact-finding mechanisms in individual countries (Regling and Watson, 2010). Even though top policymakers perceived the value of fact-finding policies, the knowledge did not travel from the national to the transnational level.

Ideational theories are ambiguous when it comes to the direction and the causality of learning. They can explain ex post facto what lessons have been learned (successful learning), but they are less useful in explaining why other ideas could have emerged but did not. Was it the impact of prevalent beliefs? Or other agents (individuals)? Or do certain mechanisms/institutions have the power to shape these ‘norms’? If ideas are responsible for both continuity and change, but ideational change is explained only ex post facto, how can we establish that it is the power of ideas that precludes other lessons from being drawn? In short, it is unclear whether ideas are only epiphenomena of broader factors in play. Perhaps they merely reflect reality instead of shaping reality.

Conclusion

In this article, we addressed an underexplored question: why do some international organisations remain resistant to drawing useful lessons from failure? The Eurozone crisis provides an opportunity to test the ability of existing theories of learning to answer the question. The crisis posed unprecedented challenges to the European integration project, and the entire Eurozone was close to collapse. European citizens quite rightly wanted to know what went wrong, who was responsible, and what should be done to fix things. Some EU member states established truth/inquiry commissions to document the causes and attribute responsibility. As the EU emphasises accountability, transparency, and legitimacy, it was highly likely for the EU to learn from the good practices of its member states and establish an independent commission of inquiry to document its own institutional and policy failures and find ways to protect the EU from a future crisis. However, the EU institutions neither adopted commissions of inquiry nor endorsed them for other crisis-ridden countries.

We argued that the EU’s fragmented organisational structure explains its partial-learning stance. In the absence of a self-reflexive organisational capacity (or a mechanism of ‘learning how to learn’), the scope of useful or relevant lessons remained limited to the effective management of the crisis, precluding good practices in other areas. It seems difficult, if not impossible, to expect policymakers to disengage from reactive learning.

Insights from theories of organisational learning help us identify the gaps in the institutional architecture of the EU affecting its capacity to learn. ‘Intelligent’ organisations have the capacity to be self-reflective, learn from their failures, and make changes to improve their performance and legitimacy. This has significant policy implications. As the post-2008 experience shows, in times of crisis, demagogues take advantage of popular discontent to score political points, in this case casting doubt on the EU’s democratic legitimacy and reducing trust in its institutions. Iceland and Ireland set up truth recovery mechanisms to uncover the causes of the crisis and offer recommendations for institutional reforms – tellingly, both countries remained immune from populism. Their experience suggests a truth commission could have greatly aided the EU’s efforts to restore trust and improve its democratic legitimacy. As we pointed out, the EU lacks an embedded self-reflective capacity. Ironically, organisations without this capacity – like the EU – could most benefit from a fact-finding body, yet they are the least likely to set one up in the first place.

Of course, truth commissions are not the only pathway to learning. The lessons drawn from the economic crisis, even though partial, were clearly instrumental in the management of the COVID-19 crisis. It would not be too far-fetched to argue that the combined incremental learning from the failures of the Troika programmes, the European Semester criteria, and the responses to the Eurozone crisis led to a different crisis management plan, one prioritising solidarity across nations and flexibility of strict fiscal rules. To be clear, the contagious nature of the crisis probably made solidarity easier, but in the absence of lessons learned from past failures, it is questionable whether the policy responses would have been as fast or as useful as they were. Even so, lessons drawn remained fragmented, ad hoc, and partial, inhibiting a more comprehensive form of learning at the EU level and preventing the best possible response to a new crisis.