Tug of War over Financial Assistance: Which Way Forward for Eurozone Stability Mechanisms?

This article analyses the development of financial assistance in the Eurozone since 2010. It argues that reforms to instruments and bodies, notably the European Financial Stability Facility, the European Stability Mechanism, and the current Covid-19 recovery fund, are best explained by a re-occurring pattern of negotiations between potential creditors and debtors based on common Eurozone interests and national cost-benefit considerations. Building on a liberal intergovern-mentalist approach, this article shows how this pattern influenced the step-by-step reform of financial assistance in the Eurozone. The threat to Eurozone stability served as a constant factor encouraging Member States to expand and deepen the assistance formula. Creditors’ cost-benefit considerations were key for retaining disincentives, a limited liability for common debt, and intermediary borrowing and lending within the financing design. However, on the back of common Eurozone interests, debtors were able to push for an increase in assistance, an expansion of assistance into areas of banking sector support, and a softening of moral hazard elements in the more recent Covid-19 pandemic. Due to creditors’ continuous insistence on safeguards and limited burden-sharing, reform outcomes were repeatedly unable to resolve the difficulties at hand.


Introduction
Since the beginning of the Euro crisis, extraordinary financial support in the Eurozone has been institutionalised through new tools and bodies with significant firepower to assist Member States in difficulties.After a first leap into this new area of Eurozone support with temporary assistance tools, Member States continued to adjust the structure repeatedly through new rules and practices and advanced deeper into the area of financial assistance introducing a permanent Eurozone support mechanism and instruments for bank recapitalisation, often subject to intense negotiations between core EU countries.
Scholars have become interested in the negotiation process finding explanatory merit in the strategic interaction between two ideational camps representing export and demand-led states (Hall, 2014), in hard intergovernmental bargaining and brinkmanship between creditors and debtors (Schimmelfennig, 2015), in the shift away from market policy to the state's core interest (Genschel & Jachtenfuchs, 2018), or in the previous institutional setting framing the choices at hand at the peak of the crisis (Verdun, 2015).Puetter (2012) acknowledged that the decision to govern new eras of EU activity was as much driven by national interests as by the willingness to find common solutions for common problems.These studies often account for complementary explanations but emphasise different explanatory variables in the intergovernmental process.Whereas strategic constructivists find that the cause of action at EU-level in the governments' rationale is based on the ideational foundation of their respective political economies (e.g., Hall, 2014;Schäfer, 2016), more rational approaches emphasise the aggregate welfare cost-benefit consideration by countries on the European level (e.g., Moravcsik, 2018;Schimmelfennig, 2015Schimmelfennig, , 2018)).
Scholars find that the post-crisis governance structure has put the adjustment costs predominantly on the economies of the countries receiving assistance with a strong emphasis on austerity being the fruit of ordoliberal German politics (e.g., Blyth, 2013;Matthijs, 2016).Scholars have emphasised the emergence of new formats and intergovernmental solutions in the aftermath of the crisis (Bickerton, Hodson, & Puetter, 2015), however, less attention has been given to the actual design of financial assistance and its adjustment in the last decade.This article complements these studies by emphasizing the concrete adjustments occurring to the financial assistance formula and the role Member States played in its continuous development.
The research question set for this article is: How can we best understand the last decade of Eurozone financial assistance reforms?As extraordinary financial assistance has the potential to carry significant costs for participating countries, the stakes in this policy are high, justifying an approach centred around Member States' interests and cost-benefit considerations in the design of support.This approach distinguishes between the potential roles of Member States in the different forms of financial assistance mostly oriented along a creditordebtor divide.Whereas Germany and other potential creditors repeatedly favoured an assistance formula with limited liabilities, cost reduction, high disincentives, and the exclusion of bank support, France and other potential debtors favoured mutualisation of debt, low disincentives, and inclusion of bank support in the assistance.Often the result of negotiations between these two coalitions favoured the creditors' design, shifting the adjustment burden on potential receivers of assistance.However, as debtors repeatedly had difficulties adjusting, a continuous adaptation of the formula was necessary in order to achieve the common interest of Eurozone stability.This led to a tug of war between the two camps over the adjustment burden and costs of reform for each step of the way.It is argued that this mixed preferences situation of common Eurozone interests and diverging interests along the creditor-debtor divide explains best how the Eurozone reformed financial assistance.
In the following section, the theoretical framework is presented.Section 3 describes the initial stage of Eurozone financial assistance established in 2010.Sections 4, 5, and 6 present the empirical analysis on several issues related to financial assistance.Section 7 concludes.

Theoretical Framework
This article borrows from liberal intergovernmentalism by assessing the outcome of integration in the area of financial assistance based on state preferences and intergovernmental bargaining.Liberal intergovernmentalism builds on the assumptions that states are bounded rational actors (Moravcsik & Schimmelfennig, 2019), acting according to their respective preferences with the general objective of achieving the most preferred outcome in the least costly way (Knight, 2018).Moravcsik (1998) provides a three-stage framework for assessing the outcomes of intergovernmental negotiations.First, states form their preferences domestically.This process is dependent on the issue at hand and the respective constituency on the national level (Moravcsik & Schimmelfennig, 2019).The more specialised a policy and the smaller the number of stakeholders, the more representative is the state's position of a specific interest group.Conversely, in a more general policy concerning a diffused entirety of taxpayers, e.g., in macroeconomics, the state represents the interests of its economy as a whole.In the second stage, the states negotiate a common agreement.Liberal intergovernmentalism argues that this process is dependent on the relative bargaining power of states deriving from asymmetric interdependence (Moravcsik & Schimmelfennig, 2019), which favours positions of states with low incentives to compromise due to their expected lower gains from a common solution (Moravcsik, 1993;Schimmelfennig, 2018).Lastly, states find agreement in specific institutional choices.This locks the agreement reached in a more or less fixed manner and tends to favour the preferences of those countries with the greatest leverage.
Additional to national preferences, this article conceptualises common 'Eurozone' or 'EU' preferences shared by all Eurozone members.This common EU or Eurozone preference is the stability of the Eurozone and the proper functioning of the common market, which this article considers as fixed.The common interests derive from the interdependence of European economies.The danger to the integrity of the Eurozone through sovereign defaults or even the breakup of the Eurozone would have devastating effects on European economies and hence on the future stability of their respective fiscal and economic positions.This creates a situation of mixed preferences in which Member States agree to resolve the difficulties at hand, but disagree about the distribution of adjustment costs (Scharpf, 1997;Zürn, 1992).In game theory, this situation is referred to as a "coordination game with distributional conflict" (Wolf, 2002, p. 39).Scholars assessed the moment of the Euro crisis as a situation of mixed preferences, in which Member States shared the common preference of safeguarding the Euro, while in parallel following diverging preference on the concrete form of adjustment, which determined how the burden is split among the states (Moravcsik & Schimmelfennig, 2019;Schimmelfennig, 2015).In 2010, the German governing coalition's behaviour showcased the common interest as it called this common or shared preference of Eurozone stability alternativlos (unavoidable).
Hence, while keeping the underlying interdependence from common preference on safeguarding the Eurozone and the common market in mind, this work will focus on the diverging preferences in the reform process.Similar to previous studies on intergovernmental negotiations, the diverging preferences are assessed based on national cost-benefit considerations.These are dependent on the expected adjustment costs linked to the materialist burden associated with financial assistance following a creditor-debtor divide (e.g., Copelovitch, Frieden, & Walter, 2016;Schimmelfennig, 2018).In the context of financial assistance, this cost-benefit analysis is related to specific elements of assistance including the scale, the timing, the source, the form, the control of assistance and the disincentives attached to it.Finally, this article emphasises the preferences of representative core-states in the bargaining processes.
Financial assistance can be generally regarded as a favourable loan or permanent transfer of capital to a recipient.These loans or transfers bear different costs for states depending on the assistance format and the expected financial involvement.Member States are expected to have entered into assistance negotiations according to their shared preferences that some form of financial assistance structure was needed or improved and have shaped the development of the Eurozone assistance structure according to the individual cost-benefit consideration of their expected involvement.
In the following, this article will assess several reforms to the financial assistance structure since the Euro crisis using preferences, intergovernmental bargain, and the institutional outcome.

The European Financial Assistance Formula
As the international financial crisis swept over to the EU in 2008-2009, non-Eurozone countries received assistance via the EU's Balance of Payments assistance facility.An instrument that borrowed via the EU on the market and on-lend to states.For Eurozone countries, no assistance tool existed from within the EU.To avoid uncertain and potentially very high adjustment costs caused by a systemic spill-over of a Greek default (Colasanti, 2016) and its potential negative signalling effect to the market about the debt sustainability of other Eurozone countries (Schimmelfennig, 2018), the Eurozone explicitly marked their common preference of Eurozone stability and established several assistance tools (Council of the EU, 2010;European Commission, 2010).First, the Greek Loan Facility (GLF) was created as a fast ad-hoc bilateral response to the Greek crisis with €80 billion of support.In parallel Eurozone countries established the European Financial Stability Facility (EFSF), a temporary 3-year special purpose vehicle that could provide loans to Eurozone members.The agreement on the EFSF allowed for the lending of up to €440 billion and was complemented by a Council regulation on the European financial stabilisation mechanism (EFSM) that allowed the EU to use its leftover financial margin as assistance as long as the crisis prevailed (see Table 1).
This first firewall was a mix of a fast-paced response to Greece's imminent difficulties and a broader approach to a potential future system of assistance.All assistance was provided via loans, a fiscally neutral way in the medium-term.Out of foresight on the costs of bilateral assistance for other Eurozone countries, the EFSF was  2019).This structure would prevent creditors from worsening their fiscal position during a time of general economic upheaval.Whereas the EU could in general fall back on its own resources as collateral for lending through the EFSM, the Eurozone countries had to provide guarantees for EFSF loans.This was achieved via a contribution key and oversubscription (guarantees), stating the share and potential financial involvement of each country in EFSF activity.The intended assistance formula aimed at limited liabilities for creditors and for the lowest possible cost, as the EU and the EFSF would lend as an intermediary, sparing the countries the direct fiscal impact of bilateral assistance.
Whereas the common interest of saving Greece and establishing a rescue fund derived from contagion risks, the EFSF followed a creditor-centred design.In particular, German and British national preferences on a non-EU instrument for Eurozone support were the reason why the Commission's alternative of assistance exclusively provided via the EU was ruled out (Gocaj & Meunier, 2013).Decision-making for assistance was in the hands of the Member States and in the case of the EFSF and GLF entirely based on national laws to ensure legal certainty and control.For the EFSM, the Council decided by qualified majority voting; for the GLF and EFSF, Eurozone countries decided unanimously with several Member States requiring approval from their parliaments.Assistance carried policy conditionality, which was set as 'strict' including extensive austerity demands, which intended to restore market confidence.These instruments increased risk-exposure of potential creditor states, but also favoured creditors' preference of including substantial obstacles to accessing assistance and to rerouting debt in order to avoid direct costs.

The Fault in Our Assistance: EFSF Reform
This initial creditor-centred design unintentionally somewhat weakened the common interest of establishing a support structure for Eurozone stability.Shortly after the establishment of the EFSF, several issues arose as the intended signalling effect to markets failed and market tension continued to rise in the Eurozone in 2011.
First, in this context, the European Central Bank (ECB) started its Securities Market Programme through which it purchased over €200 billion worth of periphery countries' government bonds on the secondary market (ECB, 2013).These interventions were contested within the ECB's Governing Council and by several Eurozone governments (Howarth, 2012), as well as by German and Dutch members of parliament (Fontan & Howarth, 2021).The main dispute was about the ECB's balance sheet stretch, with bonds from countries under increased market pressure.The intervention by the ECB effectively lowered the funding rate for these countries without them having to undergo reforms to strengthen their economic situation, which was seen by some as the root cause of their fragile market position.This undermined to some extent the previously set strict conditionality that usually accompanied extraordinary assistance.
Second, the problematic relationship between banks and sovereigns became palpable as debtor countries required substantial funds from assistance instruments because having rescued their banks they had significantly weakened their fiscal positions (Blyth, 2013;Tooze, 2018).A large part of the loans to Ireland and Greece were used to recapitalise banks, creating the situation in which the debtor countries had to substantially increase their debt.This assistance came as conditional loans, carrying strict provisions on economic reform, as it was considered a national ex-post fiscal problem (Hadjiemmanuil, 2015).However, contagion risks made financial sector assistance a Eurozone issue.
Third, the EFSF-design had two flaws, which were linked to its legal and its guarantee structure.The oversubscriptions of the EFSF were not enough to achieve the sought-after high credit rating for the total amount of €440 billion (it only achieved €250 billion).Meanwhile, Eurostat decided that guarantees for assistance under the EFSF had to be reported as government debt by the creditors (Eurostat, 2011).The GLF and the rerouting of EFSF debt had direct fiscal implications for creditor states in the short and medium-term as their debt level rose (Bundesrechnungshof, 2019).This went against the intended assistance formula, which was supposed to avoid direct national expenditure for assistance.
Fourth, the Eurozone used favourable, yet relatively high, interest rates for assistance to Greece, Ireland, and Portugal, in order to encourage swift reform implementation and hence a rapid return to the market (Colasanti, 2016;Pisani-Ferry, 2014).However, the intention to incentivise structural reforms did not have the intended effect.Creditors and European institutions did not sufficiently consider the effect of assistance via loans and their interest rates on debt sustainability.With this background, the Eurogroup decided to reform the EFSF and "adopt further measures [to] improve the euro area's systemic capacity to resist contagion risk" (Council of the EU, 2011a).
The creditor countries, particularly Germany, preferred reforms to protect their rather stable fiscal position and shield their taxpayers from assuming potential costs from mutualised instruments.They preferred to avoid being exposed to periphery states' liabilities and not to incur costs from rescuing periphery banks.Most notably, they defended the use of policy conditionality for any form of assistance in order to avoid moral hazard.Furthermore, they favoured a legal structure with effective intermediary borrowing and lending to avoid further increases of public debt through EFSF activities.Debtor states, on the other hand, favoured an increase of the EFSF firepower, a mutualisation of debt and lower interest rates for support as their fiscal position worsened.
Italy's finance minister Tremonti called mutualisation of debt the "master solution" to resolve the crisis (Hollinger, Bryant, & Peel, 2011).France favoured ECB involvement and pushed for EFSF leveraging via the ECB and a mutualised system for bank recapitalisation (Carnegy, 2011).
In early 2011, the Eurozone agreed to use EFSF funds for primary market purchases in order to allow struggling states to maintain market access.Primary market interventions were EFSF state loans with conditionality applied through a different channel.This was intended to reduce the amount of lending necessary as recipients would partially be able to fund themselves on the market (Spiegel, 2012).This solution allowed Member States to maintain control over the scale of the liabilities they assumed via interventions in the debt markets.However, as the ECB acted in parallel on the secondary market, creditors saw risks rerouted onto their central banks' balance sheets.In Germany, the Securities Market Programme was seen as the introduction of common Eurozone debt via the backdoor of the ECB's balance sheet, which was guaranteed according to the Eurozone's capital key (Bundestag, 2011a).
To regain control of assistance and ensure proper conditionality, the German coalition government considered a secondary market instrument for the EFSF, not only as an additional supporting tool but also as a replacement for uncontrolled ECB action (Bundestag, 2011b).Germany expected that Eurozone governments would take over intervention on the secondary market with conditionality, thus disarming the ECB and ensuring government control over liabilities (Bundestag, 2011c).A grand majority of German members of parliament even passed a motion stating that there was no further need for the ECB's Securities Market Programme (Bundestag, 2011d).The message was understood in the ECB and the Securities Market Programme's successor, the Outright Monetary Transactions programme, explicitly pointed to a parallel EFSF/ESM programme with conditionality as eligibility criteria for action (ECB, 2012).Primary and secondary market interventions were introduced to regain control of targeted interventions on Eurozone debt markets and allowed the recipients, in theory, to maintain market access.
Debtor countries were more concerned with avoiding adjustment costs of recapitalising their banks as none of the assistance instruments were targeted directly at the banking sector but functioned as loans to governments.France favoured a solution of recapitalisation via the EFSF (Pidd, 2011), whereas Germany preferred national solutions ("Paris et Berlin," 2011).Both countries eventually agreed on a potential use of the EFSF as a last resort for bank recapitalisation, as the market situation deteriorated and sovereign yields spreads increased considerably, putting periphery countries' debt sustainability at risk (Bundesbank, 2011).The risk associated with a cascade of defaults of periphery countries was immense and foregrounded the common Eurozone interest of stability preceding the negotiation on recapitalisation.
Together with the increase in the effective lending capacity of the EFSF from €250 billion to 440 billion (achieved by raising the guarantees), Member States agreed on a common tool for bank recapitulation.Creditors' insistence on the established assistance formula forced potential debtors to accept an indirect recapitalisation instrument.This instrument worked as a state loan, however, its conditionality was only targeted at the financial sector reducing the stigma of a full programme (ESM, 2017).Creditors refused to take over direct liabilities for difficulties occurring in other countries' banking sectors.This increased the funds available but did not ease the burden for potential debtors.Thus, the new instrument did not effectively resolve the issue of worsened fiscal positions due to bank bailouts.
The market turmoil in 2011 and the initial emphasis on disincentives by creditors in the form of strict conditionality and favourable, yet impractical, lending rates for GLF loans, brought Greece again to the brink of default.Thus, Member States decided to lower the interest rate for Greece and to lengthen the maturity of its debt (Council of the EU, 2011b).The abolishment of impractical rates for close-to-default countries was later presented as an 'impressive display of euro area solidarity' by the ESM's managing director (ESM, 2020a).This adjustment tempered the insistence on disincentives via lending rates as the common Eurozone preference and the potential risks and losses associated with a Greek default were a much larger threat than the costs associated with longer maturities and lower lending rates.However, agreement to these relieve measures and a substantial EFSF loan to Greece came again with 'appropriate incentives to implement the programme' (Council of the EU, 2011c).This could not avoid a restructuring of Greek debt held by the private sector in 2012 (Colasanti, 2016).
The future issue of counting debt as national expenditure was resolved by defining the legal structure for the in parallel negotiated design of the ESM, the successor of the EFSF, as permanent international financial organisation allowing intermediary borrowing and lending (ESM, 2019;Eurostat, 2013).The capital structure of the ESM was based on a similar logic as the EFSF, however, the ESM included a significant share of paid-in capital, which functioned as collateral together with additional national guarantees.When in late 2012 the ESM became operational, it absorbed all functions and instruments from the EFSF and reset the effective lending capacity to €500 billion.The step towards the ESM was a rectification of the faulty design of the EFSF's legal structure and increased the direct costs attributed to assistance via paid-in capital and made the formula permanent to showcase a credible commitment.However, it significantly lowered the states' guarantees and achieved the major objective of effective intermediary borrowing and lending easing the potential creditors' immediate burden.
The choice of reforms for the EFSF shows how the resolution of the crisis was a situation of mixed preferences.Eurozone countries agreed on a way forward to resolve the crisis and to achieve the common interest of Eurozone stability by increasing the lending capacity, lowering the costs for debtors and by introducing new instruments.While both, creditors and debtors, tried to deflect adjustment costs in the detailed implementation of the assistance formula.Concessions made by Germany and other creditors to include instruments on market interventions and bank recapitalisation are in detail dominated by a creditor-centred design of the assistance formula.All instruments relied on the same structure of conditional loans to states with limited liabilities, fiscal neutrality in the medium-term, and intended intermediary borrowing and lending to avoid direct costs.

Tug of War over ESM Reform and Banking Sector Support
As the crisis progressed, scholars and experts pointed towards the different pre-conditions that led Member States to seek assistance.One of these aspects was the sovereign doom loop.Arguably, this problem was made worse via the ECB's long-term refinancing operations in 2011 and 2012 as banks in Southern Europe accessed ECB financing using government bonds as collateral to buy new government bonds (Howarth & Quaglia, 2016).Another aspect was the previous use of assistance by individual states and their banking sectors.Ireland argued that it bared disproportionate costs of rescuing banks and that the Eurozone should share the cost of the Irish bailout as it reduced contagion risks for the Eurozone (Smyth, 2011).As in 2012, Spain required substantial assistance to rescue its banks, it preferred to receive recapitalisation for its banks, rather than a government loan, as it insisted that its problems were bankmade (Minder, Kulish, & Geitner, 2012).However, creditors relied on the assistance formula of loans to states.It became apparent that in order to effectively break the doom-loop and achieve Eurozone stability, bank recovery and resolution was to be shifted away from the state (van der Kwaak & van Wijnbergen, 2017).
This was acknowledged at a Euro summit in 2012.Governments saw the advantage of direct recapitalisation and agreed on providing the ESM with the possibility of recapitalising banks directly in a first step towards Banking Union.However, while France, Italy, and Spain supported direct recapitalisation, Germany and other creditor countries remained concerned about legacy issues in periphery banks and feared that the costs of losses, due to failure of nations to reform, would be spread among the Eurozone countries.Germany also feared a disadvantage for its alternative banking sector of corporate and savings banks (Commain, 2021;Howarth & Quaglia, 2016).Debtor countries favoured this instrument as it would lower their burden for potential bailouts.Creditors insisted that the establishment was coupled to progress in setting up the Single Supervisory Mechanism to monitor banks and a liability cascade to avoid disincentives for states to clean up their banking sectors (Council of the EU, 2013a).These demands were intended to prevent already struggling banks with legacy issues from tapping into the ESM funds (Howarth & Quaglia, 2016).
Other ex-ante requirements were attached to assistance including the threat to fiscal sustainability for the country in which the banks were based, systemic relevance of the bank in question, private creditor bail-in, host Member State participation in bank recapitalisation, and the inclusion of institution-specific and potentially general economic conditionality (Council of the EU, 2013b).In 2014, the Direct Recapitalisation Instrument was added to the ESM toolkit (capped at €60 billion).However, the above-mentioned requirements, mostly due to creditors' preference to avoid the moral hazard and costs of bank bailouts, made its use less likely as debtor states still had to bear most of the adjustment costs ex-ante (Merler, 2014).The step from common assistance to sovereign states to assistance to banks came with higher risks for creditors as loans to banks were riskier than state loans (ESM, 2014).Thus, creditors insisted that in this case, the formula should include additional disincentives which had the effect that the link between sovereign states and banks was only superficially cut.
A similar dynamic of mixed preferences is visible in the case of the Single Resolution Fund (SRF; for more details see Howarth & Quaglia, 2014, 2016).France's preference was for a single European fund that intervened to resolve or recover ailing financial institutions.Potential debtors favoured a common solution, which would ease the burden on banks and sovereign states, as costs would be mutualised through a European fund.Germany's preference was to safeguard its small banks, retain decision-making in the governments' hands and have a network of purely national resolution funds (Barker, Spiegel, & Wagstyl, 2013).Again, Germany feared that its corporate and savings banks would be forced to pay for failed banks in the periphery as a common system would not provide the needed incentives for debtors to restructure their banking sectors (Howarth & Quaglia, 2014).The compromise reached with Germany only included 128 larger banks and forced other Member States to accept an intergovernmental agreement for the SRF (Spiegel, 2013).
The compromise included a transitional period of 10 years (later reduced to eight), during which the SRF would be composed of national compartments.In this phase, the SRF was intended to be gradually filled with ex-ante contributions from financial institutions paying into national resolution funds until reaching a level of 1% of covered deposits (∼€60 billion; Council of the EU, 2020a).Intervention until the end of the transitional phase would be limited to the collective contributions of the respective national compartment and the overall mutualised means available to the SRF at that moment (see Table 2).With time, the national intervention quota reduces, and the mutualised means increase (for more details see Council of the EU, 2020a).
In this period the costs of assistance would predominantly be shouldered by the national banking sector requiring support.Whereas borrowing and transfers between the national compartments could be undertaken to have sufficient funds available, the banking sector receiving support would have to reimburse these loans or transfers, shielding other states from incurring losses in their compartments.Even though the Commission was given new competences, the Council maintained the possibility to object to a mutualisation (Council of the EU, 2020a).
The compromise on the transitional period followed, in particular, the German interest in using national resolution funds, including a shareholder bail-in and a reduction in the number of institutions covered by the SRF.However, the agreement also favoured the periphery states' interest in putting their larger banks under the umbrella of the Single Resolution Mechanism.The solution is a middle ground of a purely national and purely European solution, with safeguards allowing Germany to shield its banking sector and keep control within the Council.The SRF followed the logic of the assistance formula with limited liabilities for states, disincentives coupled to its setup and national involvement, as well as lending via an intermediary.
To make the SRF operational before the transitional phase ends in 2023, the Eurozone debated a potential backstop for the SRF.The Council put forward the possibility of using the ESM in the transitional phase in order to use the SRF's full capacity before all ex-ante contributions were collected (Council of the EU, 2013c).However, the same preferences leading to a semi-European solution for the SRF also fostered a similar solution for its backstop.Member States compromised on a system of bilateral Loan Facility Agreements, allowing for national bridge financing for their respective shares according to the intended size of their national compartments.Only after all means under the liability cascade of the intergovernmental agreement were exhausted, could national credit lines be drawn which had to be reimbursed in the medium-term.This meant that governments had to provide partial bailouts in the transitional period through a loan to their own SRF compartment, which would Table 2. ESM instruments and SRF.lend-on these funds.After the transitional period, the SRF compartments are supposed to merge and share the responsibility for the entire 128 banks.Creditors only agreed to the use of the ESM as backstop after the transitional phase elapsed or sufficient progress of reducing banks' exposure to risks had been made (Visco, 2019).Eurozone countries agreed on this in the revised ESM treaty in 2019 (yet to be ratified), which allowed the ESM to provide loans to the SRF up to its target level, with a nominal cap at €68 billion (ESM, 2020b).
The SRF is the first non-state funded assistance instrument with the potential to break the doom loop, however, the assistance formula upheld in the transitional period kept the burden on the country with banking sector difficulties.The intention to free Member States from adverse effects of bailouts will only be achieved partially and gradually in this period.The solution of the transitional phase was driven by creditors'-and predominantly Germany's-preferences, which included limited national liabilities, pre-requirements in banking supervision and bail-ins for SRF interventions.The common need to free Member States from the doom-loop in order to stabilise the Eurozone allowed for a deeper integration of support mechanisms favouring the debtors' positions.The rules applied in the transitional period underlined the strong adherence to disincentives and avoidance of directs costs by creditors, as most solutions only allowed for partial mutualisation and demanded that debtor states significantly participate in interventions.

Tug of War Continued: Dealing with Covid-19
After having provided immense national stimuli to their economies in order to counter the economic effect of Covid-19 (Anderson et al., 2020), several countries, including Portugal, Ireland, Greece, Slovenia, Luxembourg, and Belgium called for action in the form of a common EU debt instrument, allowing for assistance in form of grants (Dombey, Chazan, & Brunsden, 2020).The economic argument for the common European interest put forward was that the pandemic was symmetrical and that all states, regardless of their policies, were facing difficulties.However, most creditor states preferred the ESM as a potential resolution tool, which still had more than €400 billion of its capacity on standby.
The two camps agreed in April 2020 at the Eurogroup inclusive format on a 'comprehensive economic policy response,' a mix of EU budget allocations, national guarantees for European Investment Bank activity, and adjustment to ESM use.The ESM was allowed to provide credit lines of up to 2% of Eurozone GDP (€240 billion) which had to be spent on direct or indirect healthrelated expenditure (Council of the EU, 2020b).A new loan mechanism, SURE, was introduced at the EU level, which allowed the EU to borrow and on-lend €100 billion to Member States.The only condition was that the national government expenditure on short-time work and similar schemes increased since February 2020.
As of January 2021, the Council approved assistance to 18 governments via SURE with the largest share going to Italy (€27,4 billion) and Spain (€21,3 billion).The regulation worked with voluntary national guarantees of €25 billion to ensure the full capacity with beneficial lending rates, which was considered an 'important expression of solidarity' (European Commission, 2020a).Creditors adhered to the assistance formula via intermediary loans to avoid incurring direct costs and to limit their risk exposure in time through temporary instruments.The big concession on their part was the easing of conditionality for assistance, which was intended to encourage debtors to make use of the loans.
In May 2020, as the Commission projected a record economic decline in the EU, the periphery countries, but also France, refused to rely on support via Eurozone financial assistance instruments and French president Macron referred to the ESM instruments as throwing 'fake money' at the problem (Khan & Brunsden, 2020).The issue was that loans alone did not help already highly indebted countries as their fiscal sustainability was under threat and the ESM's senior creditor status could have negative effects on market lending rates.After the Commission's forecast, a Franco-German initiative proposed a €500 billion grant-based recovery fund for the EU, raised on the markets and funded by an increase in the EU's own resources and a fair taxation of the digital economy (Présidence de la République, 2020).This U-turn from the German government was defended as necessary solidarity, given that loans would not help countries with already high debt and that economic cohesion would have been severely disrupted (Bundeskanzlerin, 2020).German members of parliament argued that it was in Germany's interest to strengthen its EU neighbours due to its strong export-led market (Bundestag, 2020).
On the other hand, the Netherlands, Denmark, Sweden, and Austria (referred to as 'the frugal four') proposed only using loans for support (Rijksoverheid, 2020).In late May, the Commission combined both in a proposed recovery fund worth €750 billion (European Commission, 2020b).The dynamic was not creditor and debtor per se but between frugal states and a Franco-German-led coalition.The outcome of a record 4-day negotiation was a middle ground between both camps with a Recovery and Resilience Facility (RRF) worth €360 billion in loans and €312,5 billion in grants (Council of the EU, 2020c).
The Covid-19 overall response (see Table 3) is a step away from the previous application of assistance through conditional loans with disincentives.Even though grants were introduced and disincentives predominantly abolished, the regulation of the RRF referred to sound economic governance as part of the ex-post eligibility criteria.Thus, creditors upheld some form of conditionality.They also ensured the minimisation of direct national costs and limited liability for mutual support.The largest share of around €1 trillion worth of intended assistance Temporary limitation followed the previous assistance formula including temporary instruments in the form of intermediary loans having no direct fiscal impact and some conditions attached.

Conclusion
By applying the theoretical premise of liberal intergovernmentalism, this article provides one possible explanatory track how the Eurozone has reformed in the area of financial assistance since 2010.The re-occurring pattern of common Eurozone interests and cost-benefit considerations of creditors and debtors led to a repeated tug of war over the detailed reforms of assistance, while both sides still tried to resolve the common difficulties at hand.These situations of mixed preferences are one way of understanding the interstate bargaining process over policies with potentially high costs for Member States.
Creditors' preferences were decisive for reforms in terms of disincentives, limited liability for common debt, and the adherence to intermediary borrowing and lending to minimise direct costs.They repeatedly favoured national safeguards and the use of loans.Control and some disincentives were held on to, which reduced the effectiveness of the assistance formula and only partially allowed for a slow de-nationalisation of assistance in the case of bank-related support.Through the enabling factor of common EU and Eurozone interests, debtors were able to push for softening of moral hazard elements and an expansion of the assistance into areas of banking sector support.The common interest was also decisive for Germany and other creditors to support grants.
This explains why financial assistance, even though increasing in size and in areas of applicability since 2010, was often accompanied by a reduced involvement for creditors, a temporary form of instruments, and reinforced disincentives for debtors.Apart from disagreement on detailed application of EU and Eurozone assistance, one should however not ignore the increased volume of assistance available since the beginning of the Euro crisis, which today stands at a total capacity of around €1,3 trillion and is at least partially permanent.While some instruments are certainly more appealing for debtors than others, assistance continued to be provided to a larger extent in the form of loans.The combination of loans and grants, as well as the general risks carried by all Member States associated with assistance, indicates the commitment to the European project and underlines the institutionalised shared European and Eurozone interests.nance' held in Luxembourg in 2020, as well as the three anonymous reviewers for their valuable comments and recommendations.

Table 1 .
Assistance instruments adopted in 2010.loans as an intermediary in the same way the EU did through its Balance of Payments assistance facility (European Stability Mechanism [ESM],