Collective Policy Learning in EU Financial Assistance: Insights from the Euro Crisis and Covid‐19

Thisarticle examinespolicy changeinthe EU’sfinancialassistanceregimethrougha collectivelearning perspective.Bydefin‐ ing a financial assistance regime as the set of rules governing the disbursement and withdrawal of funding to the mem‐ ber states in the context of crisis management, the article seeks to address the following research question: How can we explain the exact form of change in the EU’s financial assistance regime between the euro crisis and the Covid‐19 pandemic? The article finds that financial assistance in the EU moved from “intergovernmental coordination” with the European Stability Mechanism to a form of “limited supranational delegation” with the Recovery and Resilience Facility and argues that such a change is due to a collective policy‐learning process. This finding suggests that the EU tends to learn from past crisis experiences, freeing itself from established institutional constraints, only when the next crisis becomes a concrete cause for concern. However, when the next crisis strikes, the EU is indeed able to radically alter its practices based on previous policy failures.


Introduction
Since the 2010s, the EU has had to confront two largescale economic crises of a different nature, i.e., the policy-induced European sovereign debt crisis (hereinafter euro crisis) and the exogenous Covid-19 pandemic.Both crises required a decisive response by the EU in terms of financial assistance.The euro crisis was the first major economic shock the EU experienced since the establishment of the Economic and Monetary Union.When it erupted in 2009, as it was yet to face a severe economic downturn, the EU was devoid of any crisis-management instrument that could provide ailing member states with financial assistance.While the EU's response to the euro crisis was thus largely "improvised" (Van Middelaar, 2019), it also constituted a litmus test for institutional resilience and formed the basis for the EU's response to the Covid-19 pandemic.This article then raises the following question: How can we explain the exact form of change in the EU's financial assistance regime between the euro crisis and the Covid-19 pandemic?
The article conceptualises "financial assistance" as the mechanism through which the EU provides member states experiencing economic difficulties with funding to preserve or restore financial stability.It identifies a financial assistance regime as a set of "formal and informal rules, practices and bodies" governing the disbursement and withdrawal of funding to the member states in the context of crisis management (Rehm, 2022).The EU's financial assistance regime is thus operationalised as the decision-making procedure (or governance system) of the major financial instrument the EU adopts in response to a specific economic crisis.As the article aims to explain policy change in the EU's financial assistance regime between the euro crisis and the Covid-19 pandemic, it selects the European Stability Mechanism (ESM) and the Recovery and Resilience Facility (RRF) respectively as comparative case studies.
The article's analysis contributes to our understanding of what the introduction to this thematic issue conceptualises as "transfer capacity" (Woźniakowski et al., in press).Both the ESM and the RRF are indeed instances of transfer powers, whereby the two financial instruments provide assistance to member state governments in the form of either "grants" or "loans," and the governments themselves are then responsible for spending those resources based on different forms of conditionality.As such, transfer capacity is opposed to "spending capacity," which implies the EU's ability to spend directly across the Union's territory, for instance, to ensure the provision of common public goods (Woźniakowski et al., in press).The EU's transfer capacity can be financed either by means of independent resources, raised through taxation (own resources) or borrowing (common debt), or by means of non-independent resources, that is capital contributions from the member states.
To this effect, the article provides an explanation for change in the EU from an intergovernmental transfer capacity of non-independent resources (with the ESM) to a form of supranational transfer capacity of independent resources (with the RRF).Specifically, it argues that, following the pandemic outbreak, EU institutions and member state governments were collectively able to learn policy lessons about the governance of financial assistance from the management of the euro crisis, leading to a change in the EU's financial assistance regime.
The above argument has the following structure.Section 2 places the article within the relevant literature.Section 3 discusses the article's analytical framework and the research strategy adopted for the empirical analysis.Section 4 examines patterns of financial assistance in the EU's response to the euro crisis and the Covid-19 pandemic.Section 5 empirically tests the policy learning argument and discusses its relative explanatory power.The final section summarises and concludes.

Institutional Change and Policy Learning in the European Union Following Covid-19
This article builds on and seeks to contribute to two strands of the literature on EU studies.First, by examining the EU's financial response to the Covid-19 pandemic, it contributes to ongoing research on EU economic governance and institutional change in times of crisis.Second, by tracing the identification of policy failures from the euro crisis and their translation into policy lessons during the pandemic, the article contributes to policy learning studies and investigates whether and how learning has the potential to bring about policy change in the EU.
To make sense of the EU's economic governance approach to the Covid-19 crisis, research on the RRF has focused on such aspects as the unprecedented provision of "grants" financed through common debt (de la Porte & Jensen, 2021), increased economic solidarity (Genschel & Jachtenfuchs, 2021), and the establishment of fiscal capacity (F.Fabbrini, 2022).When it comes to the issue of governance change the RRF involves for EU financial assistance, however, the literature is still contested.While some agree that it largely reproduces consolidated decision-making procedures (Vanhercke & Verdun, 2022), others find in the RRF an instance of "paradigm change" (Buti & Fabbrini, 2022;S. Fabbrini & Capati, in press;Schelkle, 2021).By drawing on a comparative analysis of the governance mechanisms behind the ESM and the RRF (the major financial instruments adopted in the EU's response to the euro crisis and the Covid-19 pandemic respectively), this article sheds light on the form of change the EU's financial assistance regime has undergone following the pandemic outbreak.
The literature has found the causes behind the establishment of the RRF in the exogenous and symmetric nature of the pandemic crisis as opposed to the endogenous and asymmetric euro crisis (Buti & Papaconstantinou, 2021), political entrepreneurship by powerful EU institutions (Kassim, 2023) and member states (Becker, 2022), or national material interests (Schramm, 2023).However, while these factors can explain the innovative character of the RRF as a supranational EU instrument based on common debt, they fail to account for its governance mode.At a closer look, the RRF overcomes the ESM's governance mechanism that proved controversial in the response to the euro crisis, thus suggesting the EU has learnt from the previous financial management experience.
Although the literature on policy learning is extensive, few attempts have been made at exploring whether and how policy learning occurs in the EU and its potential to induce policy change.In examining the causes behind institutional change in the EU following the Covid-19 pandemic, existing research has either focused on learning by single institutions, like the European Central Bank (ECB; Quaglia & Verdun, 2022) and the European Commission (Mirò, 2020), or within single countries, like Germany (Schoeller & Heidebrecht, 2023).Thus, whether the EU as a whole has actually undergone a process of "collective learning" is still much underexplored.This research gap is all the more relevant in that, while learning might concern single actors, conceived of as either individual (e.g., the German chancellor) or institutional actors (e.g., the European Commission), the response to any major crisis in the EU arguably involves and depends upon a "network of responders" (Moynihan, 2009) rather than a single decision-maker.Individual learning does in fact not automatically bring about collective learning and policy change (Heikkila & Gerlak, 2013).
An exception to this is Ladi and Tsarouhas' (2020) and Radaelli's (2022) study on collective learning in the EU.Though perceptive, these works put forward broad theoretical claims on how policy learning drives European integration in times of crisis that deserve to be methodologically organised and empirically substantiated.The present article takes on this endeavour.

Analytical Framework and Research Design
For analytical purposes, this article defines policy learning as the "updating of beliefs or policies based on lived or witnessed experiences, analysis or social interaction" (Dunlop & Radaelli, 2013, p. 599).As this definition might also apply to learning by single actors or institutions, collective learning includes "the collective identification and embedding of practices and behaviours" leading to policy change (Moynihan, 2009, p. 189).While policy learning and policy change are analytically distinct (e.g., actors might learn without inducing change just as change might occur without learning), learning is understood as likely to produce change (Radaelli, 2022) and "is indicated when policy changes as the result of such a process" (Hall, 1993, p. 278).The article thus adopts a macro-level approach to policy learning (Moyson et al., 2017) and deals with "governance learning" (Challies et al., 2017), or how policy actors learn about the appropriateness of different modes of governance.Specifically, the article tests whether the EU, as a collective institutional framework based on the systemic interaction among policy actors (i.e., EU institutions and member state governments), learnt from financial assistance failures during the euro crisis in its response to the Covid-19 pandemic, leading to a change in the EU's financial assistance regime.
The EU is a breeding ground for policy learning, both across policy fields and in financial assistance specifically.First, as an ever-evolving incomplete integration process, the EU has advanced in a "failing forward dynamic" through the lowest common denominator bargains among member states between one crisis and the next (Jones et al., 2016).Such integration pattern, based on the persistence of incomplete measures to address rising policy challenges, provides repeated opportunities for learning through trial and error, dysfunctional learning, and "learning to fail" (Dunlop & Radaelli, 2016;Radaelli, 2022).Second, albeit not immune to hierarchical involutions and dominance-based dynamics-as the response to the euro crisis shows (S.Fabbrini, 2016)-EU policymaking has increasingly developed into a multi-level, anti-hierarchical institutional framework that fosters ideational innovation and entrepreneurial politics, moving towards "networked governance" (Schout, 2009).In the absence of a fixed, top-down mode of governance for dealing with rising policy issues, decision-makers can work simultaneously at different levels and in different formats, exchange views, and negotiate policy outcomes among a range of potential alternatives (Piattoni, 2009).This, in turn, inevitably increases the scope for collective learning.Third, crises are believed to be key triggers for policy learning and learning-based institutional change (Deverell, 2009).While crises do not necessarily lead to policy learning, they nonetheless stand as major "win-dows of opportunity" for learning and learning-induced change (Ladi & Tsarouhas, 2020).Although the nature of the causal relation between crisis, learning, and change remains debated, the literature agrees that the temporal sequence goes from crisis to change through policy learning.In this light, not only is learning the "possible result of the way of managing and responding to crises," but European integration as such may depend on the EU's ability to learn lessons from crises (Radaelli, 2022, p. 2).In this respect, policymakers first exchange information and build knowledge based on a crisis-management experience.They thus learn lessons from policy failures associated with crisis management.Finally, policymakers can draw on those policy lessons to devise a policy response to a crisis (May, 1992).
During the Covid-19 pandemic, policy learning in EU financial assistance is expected to be facilitated by the temporal proximity with the previous euro crisis.As a large-scale economic shock, the euro crisis constitutes the most recent precedent where financial assistance was activated within the EU.In Ladi and Tsarouhas' (2020, p. 1045) own words, "it can be claimed that this time proximity has enabled quicker and deeper learning."This is all the more so as the EU governance of financial assistance during the euro crisis resulted in a manifest policy failure, both in terms of efficiency and democratic legitimacy (Donnelly, 2021;S. Fabbrini, 2013).Crisis-management experiences associated with policy failures constitute valuable testing grounds for policy learning as policy failures can act as relevant incentives for policymakers to consider institutional change.As May (1992, p. 342) has argued, "it is reasonable to presume that acknowledgement of policy failure by the policy elites within the relevant policy domain constitutes the relevant trigger for policy reconsideration and redesign." The article builds around collective learning a "putatively explanatory narrative" (Mirò, 2020, p. 2) behind policy change in the EU's financial assistance regime and puts that narrative to a plausibility test through the identification of several "observable implications" (Beach & Pedersen, 2013).As Heikkila and Gerlak (2013) suggest, collective learning unfolds through a set of subprocesses or phases, including (a) acquisition, (b) translation, and (c) dissemination.Acquisition involves the collection of information by single individuals or groups of actors about experienced "errors" or "problems."This subprocess can be triggered by changes in opportunities from the external environment, such as those stemming from a crisis outbreak, and can help policymakers discern the need for collective action.Translation consists of the interpretation of the information acquired, aimed at "drawing lessons" for the way forward.It can substantiate policy proposals that build on past failures, thus informing collective action.As both acquisition and translation are likely to occur through group dialogue and deliberation, the two phases may happen simultaneously while remaining analytically separate.Because the acquisition and translation of knowledge by individual agents or groups do not automatically lead to collective learning, dissemination finally involves the distribution of the lessons learnt across all members (of a community or an organisation) through informal bargaining or formal negotiations.To this effect, the learning actors may have to persuade or convince others that their ideas are worth being pursued through collective action (Figure 1).
In tracing the occurrence of collective learning about the EU's financial assistance regime following the pandemic crisis, the article identifies the following observable implications.First, if acquisition took place, there will be evidence of individual or institutional decisionmakers questioning the use of the existing ESM to provide financial assistance in the renewed context of the Covid-19 pandemic, pointing to its past policy failures.Second, if translation occurred, there will be evidence of policymakers putting forward alternative solutions to secure financial assistance against the pandemic crisis, moving away from the ESM governance based on the lessons learnt.Third, if dissemination was achieved, there will be evidence of informal bargaining and/or formal negotiations among EU policymakers whereby a group of them tries to persuade others that collective action to reform the governance of financial assistance in light of the pandemic is needed, thus leading to a change in the EU's financial assistance regime.Finding empirical evidence of these indicators turns the argument into a plausible causal mechanism that deserves further assessment against alternative or complementary hypotheses.
To test whether and how learning occurred in the EU's response to the Covid-19 pandemic, this article relies on (a) primary sources of EU institutions and member state governments, (b) 10 semi-structured elite interviews with EU and government officials selected among those directly involved in the negotiations for the RRF, and (c) relevant international reports and newspaper articles for the sake of data triangulation.Interviews were conducted between March and July 2022 and lasted 40 minutes on average, ranging from 20 minutes to 80 minutes.Questions included what role the interviewee's institution played in the response to the two crises and whether and how the previous euro crisis influenced the interviewee's institution's response to the Covid-19 pandemic.
The sample was diversified based on the participants' roles and institutional affiliation to ensure the validity of the interviews.Respondents included senior and lowerlevel officials from the European Commission (n = 3), Council of the European Union (n = 4), and European Parliament (n = 1), as well as member state government officials from the French Permanent Representation (n = 1) and the German Finance Ministry (n = 1).In addition, respondents served in a number of different capacities, such as policy officer (n = 6), legal officer (n = 1), policy advisor (n = 2), and policy assistant (n = 1).To maximise the number and quality of the interviews, the respondents were granted confidentiality.Therefore, in the Supplementary File, quotes are not attributed to proper names but to letters.The interviewing process stopped when "theoretical saturation" was reached, that is "the point in data collection and analysis when new information produces little or no change to emerging findings and themes" (Tracy, 2020, p. 174).The insights collected through interviews were checked against a systematic analysis of official measures taken by EU institutions and member state governments, policy statements, and press coverage.major financial response to the crisis, the ESM is an international institution outside of the EU legal framework and comes with its own set of decision-making bodies and voting rules.Its institutional structure consists of a board of governors, a board of directors, and a managing director, and its financial capacity derives from the members' capital contributions in accordance with their GDP (Treaty Establishing the ESM, 2012).

The EU's Financial Assistance
In full swing during the euro crisis, the ESM provides stability support to ESM members based on strict conditionality in the form of macroeconomic adjustment programmes.To this end, an ESM member may send a request to the chairperson of the board of governors, who may in turn entrust the European Commission and the ECB with assessing both the existence of a risk to the financial stability of the eurozone or its member states and the sustainability of public debt in the ESM member concerned.Based on such assessment, the board of governors may decide to activate a financial assistance facility in support of the ESM member.In that case, the board of governors mandates the Commission, along with the ECB and the IMF, with negotiating the conditionality scheme of the financial assistance facility in a memorandum of understanding whose terms reflect the severity of the weakness to be addressed.The memorandum of understanding needs to be approved by the board of governors and signed by the Commission on behalf of the ESM.At the same time, on a proposal from the managing director and after consent of the board of governors, the board of directors approves a financial assistance facility agreement, including the financial terms and conditions of the programme and the disbursement of financial assistance.Finally, the European Commission, along with the ECB and IMF, monitors the compliance of the ESM member with the conditionality agreed in the memorandum of understanding (Treaty Establishing the ESM, 2012, Arts.12-13).
Overall, the decision-making process for granting stability support and the disbursement of financial assistance is spearheaded by the board of governors and concluded by the board of directors, while the prevailing logic is based on mutual agreement, consensus, and unanimity.Indeed, although the board of directors approves financial assistance facility agreements by qualified majority voting (QMV), it is the board of governors that initiates and steers the decision-making process for providing stability support, and it does so by mutual agreement.This arguably makes the ESM an instrument based on the intergovernmental coordination between member state governments (Smeets et al., 2019).Based on this thematic issue's conceptual framework (Woźniakowski et al., in press), the ESM thus establishes an intergovernmental transfer capacity of non-independent resources.
While the ESM remains operational, the EU's major financial reaction to the Covid-19 pandemic consisted of the adoption of the RRF within the Next Generation EU package.The RRF is an EU treaty-based instrument and stands as the core programme of Next Generation EU, which is legally integrated into the 2021-2027 Multiannual Financial Framework.Its financial capacity derives from the unprecedented large-scale emission of common debt through the European Commission's borrowing operations on the financial markets and from an increase in the Union's own resources (Regulation of the European Parliament and of the Council of 12 February 2021, 2021).Two decision-making procedures arise from the RRF-one for the disbursement of financial assistance and the other for the suspension (and lifting thereof) of financial payments.Both procedures revolve around the European Commission and the Council, but the balance of power tilts towards the Council in the former procedure (disbursement) and towards the Commission in the latter (suspension and lifting of suspension; S. Fabbrini & Capati, in press).
In practice, the Commission assesses member states' national recovery and resilience plans (NRRPs) based on a specific list of criteria.On a proposal from the Commission, the Council approves such an assessment by QMV, paving the way for the Commission's decision on the disbursement of the financial contribution.An emergency brake allows member states to exceptionally ask the president of the European Council to bring any NRRPs to the next European Council meeting for discussion, in which case the Commission cannot authorise the disbursement of the financial contribution until the European Council has discussed the matter.The powers of the European Council on NRRPs are, however, limited in both time and scope.On the one hand, the whole process should not take longer than three months since the Commission first asked for the opinion of the Economic and Financial Committee.On the other, member state governments have no veto power over the disbursement of financial contributions, and the final decision on authorising such disbursement lies with the European Commission.The European Commission can propose to the Council to suspend all or part of the financial assistance under the RRF or to lift such suspension, with the Council acting by reversed QMV.This slightly diminishes the decision-making role of the Council, as it needs a qualified majority to reverse the Commission proposal.
The institutions involved in the decision-making process and their voting rules suggest the governance of the RRF is not fully supranational and by far not intergovernmental.A fully supranational procedure would entail the Council and European Parliament sharing decisionmaking powers on a Commission proposal, with the Council acting by QMV and the Parliament by a simple or absolute majority (as per Art.294 TFEU).Under the RRF, the Commission has the monopoly of policy initiative, while the Council decides on a Commission proposal alone.At the same time, intergovernmental governance would imply a preeminent role of the European Council and the Council, both acting by unanimity (as per Art.24 TEU).In this case, the European Council is only allowed to discuss an NRRP before the Commission can authorise the payment.Moreover, member state governments within the Council and, even more so, within the European Council can exercise no veto power at all.Hence, the governance of the RRF constitutes a form of "limited supranational delegation."Contrary to the ESM, the RRF thus establishes a form of supranationallike transfer capacity based on independent resources.Table 1 below summarises the governance of the RRF in comparison with that of the ESM.
What emerges is that the EU's financial assistance regime shifted from intergovernmental coordination as epitomised by the ESM in response to the euro crisis to a form of limited supranational delegation as epitomised by the RRF in response to the Covid-19 pandemic.The EU thus moved from an intergovernmental transfer capacity of non-independent resources to a form of supranational-like transfer capacity of independent resources.

Different Outcomes for Different Crises: Collective Learning in EU Financial Assistance
What explains such a change in the EU's financial assistance regime between the two crises?This section empirically tests the plausibility of the policy learning argument in three steps.First, it discusses the policy failures associated with the EU's financial management of the euro crisis through the ESM as identified by policymakers and the epistemic community.Second, it traces the occurrence of collective learning by examining the mechanisms of acquisition, translation, and dissemination of the relevant policy lessons and how such mechanisms led to the adoption of the RRF as an instrument of "limited supranational delegation."Third and finally, the sec-tion reflects on the results of the empirical analysis and discusses the relative explanatory power of policy learning compared to alternative hypotheses.

Policy Failures
When the Covid-19 pandemic broke out in March 2020, the ESM was the single major crisis-resolution tool in the EU.It thus stood as a "default option" for eurozone countries in need of financial assistance (Howarth & Quaglia, 2021, p. 7).In fact, prospects of relying on the ESM as the major response to the pandemic crisis were still prevalent in EU circles at least until early April 2020 (Bufacchi, 2020).However, no eurozone member opted for activating the instrument in their response to Covid-19, not even when its new health-related conditionality-light credit line (i.e., the Pandemic Crisis Support) became operational in May.After all, the ESM had come out of the experience of the sovereign debt crisis scratched and ailing.Owing much to its intergovernmental logic, the ESM had failed adequately to meet the criteria of efficiency and legitimacy in dealing with the crisis (Donnelly, 2021;S. Fabbrini, 2013).
In terms of efficiency, unanimity rules in the ESM decision-making system created multiple veto players, each virtually able to stop the adoption of any solution to the crisis.This allegedly contributed to slowing down the EU's reaction to the financial turmoil and made it difficult to stop the spread of the crisis from Greece to other Southern European member states (Interviews E and I).In terms of legitimacy, and again due to their intergovernmental character, decisions in the ESM were taken with little (if any) consideration for the European Parliament and national parliaments, despite having  Howarth & Spendzharova, 2019).On top of that, the asymmetric vulnerabilities of eurozone members to the euro crisis and the ensuing divide between "creditor" and "debtor" countries allowed the most powerful actors at the time-notably Germanyto "weaponize" the ESM as a way of imposing "practical authority over other institutions, core EU policy principles, programmes, institutions and regulations and [placing] conditions on other countries" (Donnelly, 2021(Donnelly, , p. 1576)).Over time, this greatly contributed to increasing levels of public distrust towards the EU in general and its financial assistance practices in particular (Interviews G and J; Schmidt, 2020).

Acquisition
The shortcomings in the EU's financial response to the euro crisis provided EU policymakers with relevant hints on how (not) to go about financial crisis management during the Covid-19 pandemic.When the pandemic broke out, the experience of the euro crisis was still very vivid to EU policymakers, as were the policy failures in the EU's response to it.As one EU officer admitted: Even if [most] leaders between the two crises changed, governments and EU institutions have a living memory and especially with respect to the use of the ESM they realised what the huge repercussions of how they dealt with the previous crisis were.(Interview C) At an early Eurogroup meeting on 16 March, Italian Prime Minister Giuseppe Conte was among the first to claim the inadequacy of existing financial tools, stressing that "the ESM was crafted with a different type of crisis in mind" and that "probably the only way forward would be the creation of a common European debt instrument" (Johnson et al., 2020).Conte's concerns were echoed on 25 March by a French-led initiative (Interview F) including nine member state governments who, acting on the basis of "past experiences" and "thorough exchange of information," called on a "common debt instrument issued by a European institution" to counter the damage caused by the pandemic, thereby dismissing the ESM (Wilmès et al., 2020).The acquisition of knowledge based on the management of the euro crisis soon assumed a collective character when, in their joint statement of 26 March, the members of the European Council stressed the importance of "drawing all lessons from the crisis," concluding that in "that respect, the time has come to put into place a more ambitious and wide-ranging crisis management system within the EU" (European Council, 2020a, p. 6).
While the exact governance features of the new financial instrument were yet to be discussed at this stage, it was already clear that, due to the manifest pol-icy failures in the EU's response to the previous crisis, "the new system would [have to] be much more supranational in comparison with the ESM" (Interview A) and that "something was learnt with respect to financial governance and how the ESM fared in its management of the euro crisis" (Interview D).At this time, the idea of relying on the ESM as the major tool to address the Covid-19 crisis had completely vanished as it was perceived as "poisonous" by the policymakers' large majority (Interview B).
The acquisition of knowledge from the financial response to the euro crisis by some key actorsincluding the Italian prime minister, the French president, and other government representatives, mostly from Southern Europe-thus paved the way for the subsequent translation of it into policy proposals for a new financial instrument to address the Covid-19 pandemic.

Translation
In the conclusions to their meeting of 26 March, government heads had invited the European Commission to come up with proposals for Europe's recovery.On 16 April, in a speech at the European Parliament, European Commission President von der Leyen thus put forward the idea of an ambitious "Marshall Plan for Europe's recovery."She acted on the premise that "Europe has had economic crises before" and that "the moment has arrived in which we must know how to discard old burdens," adding that "this is the lesson we need to learn from this crisis" (von der Leyen, 2020).On the same day, the European Policy Centre published a discussion paper identifying several "key lessons [that] can be learned and applied from [the euro crisis], including the need to 'jointly set up and finance a common Covid-19 recovery and growth fund'" (Emmanoulidis & Zuleeg, 2020, p. 3).One month later, French President Macron and German Chancellor Merkel came up with their joint initiative for a common debt instrument to replace the ESM in addressing the pandemic, one based on "an in-depth reflection on the lessons we need to draw" (Présidence de la République Française, 2020).Reporting on it, the Financial Times acknowledged that: The lesson of past crises is that inadequate measures sharpen disagreements among governments, stimulate public frustration with the EU and sow doubts in financial markets about the eurozone's stability.The French-German initiative stands out from crisisfighting measures deployed in the sovereign debt and bank turmoil of a decade ago.("Franco-German rescue plan is a big step forward," 2020) Along these lines, in its legislative initiative for the establishment of the RRF of 28 May, the European Commission advanced a largely supranational, comitology-like governance limiting the Council's role to the suspension of payments on a recommendation from the Commission and based on the use of reversed QMV rather than unanimity.The legislative proposal thus distanced the governance of the RRF from that of the ESM.The Commission then emphasised among the "grounds for the proposal/initiative" exactly the "lessons learned from similar experiences in the past" (European Commission, 2020, p. 34).In that respect, an EU officer revealed that "the very negative experiences from the ESM bailouts in Portugal, Spain, Ireland and Greece were contemplated and contributed to the greater role by the European Commission in the definition and governance of the RRF" (Interview G).Following the Commission's proposal, on 9 June, the German, Portuguese, and Slovenian governments presented the 18-month programme of their Council presidency, suggesting they would steer upcoming negotiations by "drawing all lessons from the crisis and tackling its socio-economic consequences" (Council of the European Union, 2020).
In this phase, some EU and national policymakersnotably the Commission president, the French president, and the German chancellor-translated the lessons learnt from the mismanagement of the euro crisis into policy proposals for establishing the RRF around a supranational governance system that differed from the intergovernmental ESM.This opened a process of dissemination of new ideas through hard bargaining and negotiations, leading to collective learning.

Dissemination
The Commission's initiative was followed by several rounds of negotiations before an agreement could be reached.While learning through bargaining might sound odd, negotiations can produce information and shed light on alternative courses of action which would otherwise remain uncharted (Dunlop & Radaelli, 2016).In particular, the governance of the RRF became "the single most important and difficult question" that the political leaders would deal with (Interview F; Ludlow, 2020).While a large majority of policymakers-the so-called "solidarity coalition" (S.Fabbrini, 2023)endorsed the RRF's governance mechanism as per the Commission's scheme, a small coalition of veto players-the self-defined "Frugal Four," including Austria, Denmark, the Netherlands, and Sweden-opposed it, favouring unanimity in the Council instead.
Upon assuming the Council presidency on 1 July, the German government thus circulated a draft proposal providing that the Council would not only suspend payments on a recommendation from the Commission but that it would have a say on any phase of the process and approve the Commission's assessment of NRRPs by QMV (Ludlow, 2020).The German draft was debated at the EU ambassadors meeting on 8 July.On that occasion, Dutch EU Permanent Representative De Groot appreciated Germany's effort but said the Netherlands still favoured unanimity voting in the Council on a Commission recommendation.As an insider argued, by then "the Frugals themselves had become increasingly aware that a solution like the ESM would be impracticable for the Covid-19 pandemic and only pushed for unanimity to obtain a greater role of governments in the Council" (Interview H).Overall, therefore, the German proposal was hailed as a big progress in the negotiations by the Frugal Four as it somewhat moved the balance of decision-making powers under the RRF from the European Commission to the Council ("POLITICO Brussels Playbook: Michel's not taking 'no,'" 2020).
On 10 July, in his "negotiating box" ahead of the European Council meeting of 17-21 July, Charles Michel reiterated that "it is essential to learn the lessons" with a clear reference to the unanimity issue (European Council, 2020b), and supported the German blueprint for the governance of the RRF.The European Commission's Representative Gert-Jan Koopman welcomed it and said that "the Commission was not opposed in principle to enlarging the Council's role" in the governance of the RRF (Ludlow, 2020, p. 28).Government representatives of the solidarity coalition appreciated the preservation of an overarching supranational system of financial assistance, while the Frugal Four started softening their positions.It was on this basis that a compromise on the governance of the RRF was achieved at the European Council meeting of 17-21 July.The Dutch government insisted that the member states should have continued control over the national recovery plans, claiming for them the power to stop the activation of financial assistance in case an NRRP appeared not to be in line with the established criteria.Such a request was opposed by both the Italian government and the Commission, who feared this could jeopardise the supranational structure of the recovery instrument (Ludlow, 2020).
Working closely with Merkel, Michel thus put forward a clause providing that, in case of doubts or concerns, the member states could ask to discuss any NRRP at the next European Council meeting before the Commission could recommend the activation of financial assistance.At the same time, the European Council would have no veto powers over the disbursement of payments and the last say would continue to lie with the Commission (European Council, 2020c).In this way, they were able to strike a deal with the Frugal Four without shaking the supranational nature of the RRF's governance.The added clause, known as the "emergency brake," represented the fundamental compromise behind the recovery instrument (Interviews F, H, and J) and allowed the establishment of the RRF around a form of "limited supranational delegation." Pointing to collective learning, the final RRF regulation, published on 12 February 2021, reported: The Facility should be a dedicated instrument designed to tackle to adverse effects and consequences of the Covid-19 crisis in the Union.It should be comprehensive and should benefit from the experience gained by the Commission and the member states from the use of other instruments and programmes.(Regulation of the European Parliament and of the Council of 12 February 2021, 2021, p. 5, emphasis added)

Discussion of Results
Overall, in the interviews with policymakers involved in the EU's response to the pandemic, learning from past crisis experiences was identified as one of the main factors behind the establishment of the RRF by all but one (n = 10).In particular, policy learning emerged from the interviews as one of three competing-but not mutually exclusive-narratives on the causes behind governance change in the EU's financial assistance regime following Covid-19, with the other narratives revolving around the nature of the pandemic crisis as "exogenous" rather than "endogenous" (as was instead the euro crisis) and the effects of the pandemic crisis as "partly symmetric" rather than "fully asymmetric" (as were those of the euro crisis).A minor narrative, which a few interviewees also mentioned as a potential cause of change in the governance of financial assistance, concerned Brexit and the constraining effects of the "British veto" in the past (Interviews A, D, and F).
As for the relative explanatory weight of policy learning in relation to the other narratives, no unanimous view emerged from the interviews.While some posited that policy learning was conditional upon the different nature and effects of the Covid-19 pandemic compared to the euro crisis (Interviews A, B, and E), others argued that the experience of the euro crisis would have urged EU policymakers to learn key policy lessons anyway (Interviews C, D, and G).For our purposes, however, policy learning has an analytical role of its own.While the different nature and effects of the pandemic vis-à-vis the euro crisis may indeed shed light on policymakers' willingness to set up a dedicated financial assistance mechanism larger in size and more comprehensive in scope than the ESM, it is specifically through learning from the policy failures of intergovernmental coordination that change in the governance of financial assistance towards a form of supranational delegation can best be explained.

Conclusion
This article has examined policy change in the EU's financial assistance regime between the euro crisis and the Covid-19 pandemic.It has shown that the governance of financial assistance in the EU moved from intergovernmental coordination with the ESM as a response to the euro crisis to a form of limited supranational delegation with the RRF in response to the Covid-19 pandemic.By relying on official documents, semi-structured elite interviews, and international reports, the article has argued that such a change was due to a process of collective learning.To do so, it has traced how the outbreak of the pandemic crisis prompted the acquisition, trans-lation, and dissemination by EU and national policymakers of policy lessons from the management of the euro crisis and the use of the ESM.It has then shown how the unfolding of such a process ultimately led to a governance change in EU financial assistance with the establishment of the RRF.
The article makes both a theoretical and an empirical contribution.Theoretically, it applies the concept of policy learning to crisis-induced institutional change in the EU.In particular, it turns collective learning into a testable causal mechanism behind policy change in EU financial assistance following the outbreak of Covid-19, examining its plausibility and discussing its explanatory power compared to alternative hypotheses.In doing so, the article corroborates the potential of a policy learning framework to account for European supranational integration in times of emergency politics.Empirically, the article sheds light on the policymaking dynamics leading to the establishment of the RRF, focusing on the role of actors and their motivations.It shows that EU collective action to address the pandemic crisis was informed by the policy failures of the ESM during the previous euro crisis and that policy learning led to a change in the EU's financial assistance regime through hard bargaining and negotiations.
The article's findings raise two points of discussion.First, despite the failures of intergovernmentalism in the EU's response to the euro crisis, the ESM was in full swing up until the coronavirus outbreak, marking a long period of institutional path dependence.Intergovernmental coordination was only challenged after Covid-19 had turned into a global pandemic, forcing EU and government officials to come to terms with the mismanagement of the previous economic crisis.This may suggest that the EU tends to learn from past failures, freeing itself from established institutional constraints, only when the next crisis provides a window of opportunity for institutional change.Second, as the process of collective learning indicates, when the next crisis strikes, the EU is able to radically alter its governance methods based on previous policy failures.This may imply that European integration proceeds mostly through "critical junctures" leading to radical and abrupt changes, rather than gradually or incrementally between one crisis and the next.
the Post-Covid EU, held at LUISS University (Rome) on 17-18 March 2023, and the three anonymous reviewers for their helpful comments.

Figure 1 .
Figure 1.Visual representation of the "collective learning" process behind policy change in the EU's financial assistance regime.

Table 1 .
The ESM and RRF: Governance and capacity-building.