Capitalising on External Constraint: Six Things You Should Know About Eurozone Bailouts

In the last decade, five Eurozone governments in economic difficulty received assistance from international lenders on condition that certain policies specified in the Memoranda of Understanding were implemented. What room of manoeuvre did the governments of these countries have? After conditionality, to what extent were governments willing and able to roll back changes imposed on them by the international lenders? Do we find variation across governments, and if so, why? <br>This paper addresses those questions, summarizing the main findings of our book (Moury et al. 2021) on constraints on national executives in the five bailed out countries of the Eurozone during and beyond the crisis (2008-2019). We show that, despite international market pressure and creditors’ conditionality, governments had some room for manoeuvre during a bail out and were able to advocate, resist, shape or roll back some of the policies demanded by external actors. Under certain circumstances, domestic actors were also able to exploit the constraint of conditionality to their own advantage. The paper additionally shows that after a bail-out programme governments could use their discretion to revert the measures which bring the greatest benefits at a lower cost.


Introduction
In the last decade, five Eurozone governments received financial assistance from the International Monetary Fund (IMF) and the European Union (EU), on the condition that a long list of policies were implemented. This is often described as a period in which powerless bailed out governments had no choice but to follow the lenders' prescriptions and the effectiveness and fairness of conditionality has also been put into question.
Despite the widespread media coverage, leaks and memoirs, there is actually very little empirical research about how domestic governments have dealt with conditionality. While the crisis has attracted a lot of academic attention, most of the studies have examined the content and effects of the reforms passed during the crisis (see for example Matsaganis 2018, 2019). Many have also stressed and enlightened the shift in power at the European level to the benefit of richer member states like Germany (see Finke and Bailer 2019;Schimmelfennig, 2015;Tsebelis 2016) and of the European institutions (Bauer & Becker 2014;Schmidt 2016;Afonso 2019). Other studies have focused on the reasons why European leaders chose to include the International Monetary Fund (Henning 2017) and the subsequent divergence of views between the lenders (Clift 2018;Lütz and Kranke 2014). Finally, others have looked at the crisis from an international political economy perspective, stressing the centrality of domestic interests and interstate bargaining in shaping the European decisions (see for example the special issue from Copelovitch, Frieden and Walter 2016).
This large bulk of research, however, often assumes that Eurozone bailed out countries were forced to implement very specific reforms against their will. We believe that this assumption is not fully correct. Together with a handful of others (for example Hick 2018 for Eurozone bail-outs; Vreeland 2004 andPop-Eleches 2008ab for IMF lendings), we claim that bailed out governments had room for manoeuvre and, under certain circumstances, were able to exploit the constraint of conditionality to their own advantage.
Our understanding is even more limited in relation to what happened when conditionality finished, and to the policy-makers' leeway and motivations to revert (or alternatively maintain) changes introduced under conditionality. Indeed, with the exception of studies on compliance after Eastern enlargement (for example, Epstein and Sedelmeier 2008;Levitz and Pop-Eleches 2010) and on the longterm economic effects of IMF programmes (i.e. Remmer 1986;Vreeland 2004), scholars have made surprisingly few systematic attempts to measure the resilience of reforms adopted under conditionality. As an exception, Rickard and Caraway (2019) observe that the public sector wage cuts induced by IMF conditionality do not persist in the longer-term; and the special issue edited by Moury and Afonso (2019) looks at the (relatively few) policy reversals in Spain and Portugal after the Eurozone bail-outs. We argue that after a bail-out programme governments could use their discretion to revert the measures which bring the greatest benefits at a lower cost.
In this working paper, we summarize the findings of a large investigation that we pursued in the five Eurozone bailed out countries (Moury et al. 2021). We start by briefly presenting our arguments and methodology choices before describing our main findings and concluding.

On the Constraints faced by European governments
As several authors have noted, parties in governments are limited in their capacity to respond to citizens' demands, Mair 2009;Laffan 2014). To start with, governments that increase their deficit and debt excessively, or deviate too far from supply-side policies in this context risk damaging their currency's Robert Schuman Centre for Advanced Studies Working Papers credibility and their firms' competitiveness (Schmidt, 2002:22-23;Clift and Rynner 2014;Hallerberg and Basinger 1998). Governments in the EU are also bound by the European treaties (and the rules thereof) which, unlike other constitutions, are not only extensive but difficult to amend (De Witte 2009).
The shrinking discretion of European governments became crystal clear during the sovereign debt crisis (Streeck 2014;Ladi and Graziano 2014;Ferrera 2017). European economic governance rules were modified over this period, and the new instruments increased the EU's surveillance and enforcement capacity regarding the implementation of structural reforms. A fundamental constraint on European governments during the crisis also came from the increasing divergence between the costs of Eurozone countries' sovereign bonds. Periphery countries became increasingly concerned about their credibility and therefore engaged in spending cuts and reforms (Moury and Standring 2017).
These circumstances forced some EU countries to respond to implicit and explicit conditionality. Implicit conditionality, which is 'based on an implicit understanding of the stakes and sanctions involved, underlain by some measure of power asymmetry' (Sacchi 2018, p. 1) was exercised, for example, by the ECB when it warned Italy and Spain to stop buying government bonds unless they pursued reforms (Bosco and Verney 2012:138;Lutz et al 2020;Sacchi 2018). Implicit conditionality was also enforced by EU actors and/or powerful member states in 'backroom' diplomacy' (De la Porte and Natali 2014:735) -their influence deriving from the liquidity they could eventually provide to weaker countries and/or from the knock-on effect their declarations would have on investors (Moury, De Giorgi and Pita Barros 2020). Countries that asked for a bail-out -such as Greece, Ireland, Portugal, Spain and Cyprus -were additionally subject to explicit conditionality as the progressive payments of the loans granted to them were conditional upon a series of reforms detailed in Memoranda of Understanding (MoU).

On the Leverage of Executives during a bail-out
Our main argument is that, despite those constraints, bailed out executives had some leeway in both the timing, design and implementation of the MoU, a leverage that could be explained by two principal factors. First, because 'ownership', i.e. a country's interest 'in pursuing reforms independently of any incentives provided by multilateral lenders' (Drazen 2002:37), is highly valued by international lenders. This is because programme ownership raises the probability of the programmes' success and hence avoids a waste of resources (Drazen 2002, Khan andSharma 2001: 28).
Second, domestic leverage emerges from the fact that international lenders need insider knowledge to design efficient plans. As one IMF official explained: "It's not like the troika comes with a list of things and says that they have to do these things and then get the money. It doesn't make sense partly because only the government knows how things work "(quoted by Moury and Freire 2013:46).

On the crisis as a window of opportunity for reformist ministers
Ownership, however, is not without ambiguity, since conditionality would not be necessary if the country was willing to undertake the policies in the programme even in the absence of a bail-out. Our argument here is that the bail-out is sometimes useful for executives to pass measures that they favour but could not have passed in the past.
It is useful to distinguish between three scenarios here (Bird and Willett 2004). One is that the government radically disagrees with the principles and assumptions of the lenders detailed above. In this case, the bail-out seeks to coerce the government to make fundamental changes to its policies but the programme's success is not guaranteed because of this lack of ownership. A second scenario is that the executives agree entirely with the lenders about the diagnostics and the remedies to be adopted, and are willing and able to pass them whatever the costs. In this case, little is changed by conditionality.
In the third scenario, which we assume to be the norm, the government neither radically disagrees nor agrees with the lenders' principles. First, the executive might concur with some postulates and disagree with others. It would be the case, for example, of a government which agrees that rents enjoyed by some firms should be decreased but not with the proposal to cut civil servants' wages. In that case, the government can pass some of its preferred policies; but is constrained to pursue others.
Also, the executives might be divided between reformist and pro-reformists ministers, possibly from different parties, party factions or professional backgrounds (think for example of the likely conflict between a Labour Minister from a Trade Union and a Finance Minister from the Central Bank). In this case, the arrangement with the troika may 'tip the balance' in favour of those supporting economic reforms (Bird and Willett 2004:429).
Finally, in such a third scenario, the executives might consider the reforms required by the troika desirable but, outside of the programme, would not be willing or able to pay the costs associated with them. During a bail-out, these costs might be lower as opponents to policy changes take the risks of losing the loan into consideration when choosing their strategies (Vreeland 2005). Also, international negotiations usually take place behind closed doors. This grants executives privileged information about the constraints they face (Moravcsik 1994) and hence the possibility to use 'false dilemmas' (Moury and Gago 2017) -that is, a choice between two options (for example the policy change or bankruptcy) despite there being other solutions available.
Additionally, when a country is bailed out, pro-reform ministers might shield themselves, at least partially, from the electoral costs of unpopular measures. We know that a crisis decreases the electoral costs of unpopular measures (Giger and Nelson 2011;Costa Lobo and Lewis-Beck 2012).
Assuming that this third scenario is the most likely, our second argument is that as ministers have some leeway in the designing of adjustment programmes, they are able to use conditionality, at least to some extent, to pass measures that they had been favouring but could not previously have passed. In other words, and to quote Putnam, bail-out sometimes allows executives "to do what they privately wish to do, but are powerless to do domestically" (Putnam 1988: 457; see also Schelling 1960:22). This domestic empowerment of executives due to the conditionality imposed by the EU or the IMF has been demonstrated empirically. For example, Dyson and Featherstone (1996), demonstrate how a small technocratic elite in Italy chose to submit their country to the constraints associated with the economic and monetary union (EMU) in order to 'bind themselves to the mast' and impose policy changes they would not otherwise have been able to pass and maintain. Similarly, Vreeland (2005) demonstrates that executives facing more veto players are more likely to turn to the IMF.
In relation to the specific cases of the bailed out Eurozone countries, there is also some evidence of the leverage of executives. For example, Hick (2018) examines the troika's influence on the retrenchment and reform of social security in Ireland and concludes that a large part of the decisions had domestic origins (see also Dukelow 2015). Moury and Standring (2017) show that executives in Portugal were to a great extent convinced of the need to introduce spending cuts and deregulate structural reforms, and passed reforms that went beyond the lenders' original demands (see also Lutz et al. 2019;Asensio and Popic 2019). MacCarthaigh and Hardiman (2019) argued that the troika allowed the Irish government and the administrative elite to invoke the troika's force majeure to strengthen their own hand vis-à-vis the public sector workforce and to engage in a profound reform of the Irish public administration. Finally, Cioffi and Dubin (2016) show that Spain's conservative government pursued radical neoliberal labour law reforms that were not requested by external actors. While interesting, these studies focus on most likely cases such as the (centre)-right governments of Ireland, Spain and Portugal, and often cover single-policy areas. In this paper and in our book, we further enhance and deepen this argument by covering all policy areas and all bailed out governments of the Eurozone. Robert Schuman Centre for Advanced Studies Working Papers

On the importance of Partisanship
Given that bailed out governments have leverage and can use external constraints to introduce measures they politically favour, then the partisan composition of governments should make a difference even during a bail-out. This expectation is coherent with the partisan-effect theory that posits that political partisanship has an impact on socio-economic reforms during a recession or an adjustment programme. More precisely, the hypothesis holds that left-wing political parties are more favourable towards the expansion / limited retrenchment of the welfare state, the redistribution of wealth, and more protective of the labour condition. In contrast, right-wing parties are expected to strive to reduce government intervention, increase the disciplining effects of market mechanisms and limit the breadth and depth of the welfare safety net (Pop-Eleches 2008;Starke et al. 2013). Our third argument is therefore that partisanship matters and governments with different persuasions will implement different policies during a bail-out.

On the bargaining context
Of course, not all states in international negotiations have equal weight; that is why we should bear in mind that the leverage of governments -and their related capacity to pass their preferred policies during the bail-out -is mediated by their material bargaining power.
Bargaining power varies primarily in line with one's material resources (i.e. size, GDP). In the case of the recent bail-outs, scholars have shown that the debtors' leverage varies as a function of their GDP and corresponding share of ECB capital and their level of debts (Offe 2017). Amongst the borrower countries, Greece was therefore clearly in a weaker position than the others.
But other factors impact states' material bargaining power. For example, it has been proved that the credibility of the lenders' threats not to grant assistance decreases the borrowers' leverage. This credibility itself also depends on the level of exposure of the creditors' banks in the countries at risk (Howarth and Quaglia 2016;Hennessy 2017). This, for example, explains why, when a country -such as Spain -is 'too big to fail' (Wildasin 1999), is in a stronger bargaining position than smaller countries.
Fourth, many international scholars and bargaining theorists have stressed that saliency is an important factor in negotiations. Saliency refers to the importance which actors attribute to a specific issue, and hence, to their willingness to invest political resources in it (Beyers et al. 2017;Meissner and Mc Kenzie 2019). Hence, we could expect that when an issue is salient for the lenders, the government will have less room for manoeuvre. Alternatively, governments that invest a lot of political capital in an issue might eventually win the negotiations on this particular point. The counterpart of this is that the more salient an issue is for an actor, the bigger the concessions it is willing to make in other areas (Keohane and Nye 1977;Moravcsik 1998).
Fifht, credibility -which we define as the perceived intention to negotiate honestly and the appraised ability to respect one's commitments -is a crucial factor for domestic leverage (Leach and Sabatier 2005;Moury 2013). The credibility of governments depends on a variety of factors. First, trust (in this context, the perceived intention to negotiate honestly) is a prerequisite for credibility. As Nelson discusses and empirically demonstrates for the IMF, if the international lenders think that the executive can be trusted "to do the right thing", then it will leave them freer (Nelson 2014:307-308). As the trustworthiness of a given policy-maker emanates from previous interaction, political instability (governmental resignation, new elections, reshuffles, etc.) is detrimental to it (Tallberg 2008). Moreover, the partisan composition of government is also important for trust and leverage. This is based on the contention that success is more likely in negotiations when actors share the same preferences and values as their negotiators (Zahariadis 2016) and have less extreme positions (Garrett and Tsebelis 1996). The government's credibility, finally, depends on the institutional capacity of the state to implement the deals and respect the rules (Sacchi 2018; Hennessy 2017; Exadaktylos and Zahariadis 2014 1 ). Hence, the relationship between credibility and leverage is circular and self-reinforcing rather than unilateral. While, as discussed earlier, lenders might find it advantageous to give the government leverage in the negotiations, shared understanding of the problems and solutions also increases government leverage.

After Conditionality
After 2013, Eurozone bail-out programmes started to come to an end and the economic situation improved (to some varying extent). This obviously made executives freer than before. During this phase, lenders and investors did not scrutinise policy developments so closely, and declining yields made governments less dependent on the conditionality. In that context, the conditions that allowed reformists ministers to push for their favourite policies were no longer met; and the executive, even when there was continuity in office, was tempted in this new context to please its constituency or surrender to interest groups' pressure.
Yet external constraints did not disappear after the bail-outs. EU rules within the Stability and Growth Pact were tightened, at least until the pandemic, so that the EU could monitor a large spectrum of indicators (including wages, house prices, private debts, etc.) and sanctions for noncompliance were easier to pass (Bauer and Becker 2014). Countries benefiting from the ECB's quantitative easing were still vulnerable to the bank's conditionality. A third, more informal source of pressure was that recently bailed out countries wanted to avoid being seen as going back to the same risky behaviours. Such an image could diminish their bargaining power when dealing with their peers and EU institutions (Sacchi 2018); and weaken investors' trust in their sovereign bonds (Tavares 2004;Matthijs and McNamara 2015).
To sum up, governments were pulled and could go in two contradictory directions, towards reversals or the maintenance of what had been done under conditionality. The question asked was which policies they would revert. Our answer to this, and our final argument, is that governments will selectively pick the reversals that bring more benefits than costs to them.
First, we claim that policy-makers would revert policies when it is most electorally beneficial to do so. For example, we expect governments to revert the civil servants' tax rises or salary cuts rather than VAT increases (which imply diffuse benefits). In that vein, Afonso and Bulfone (2019) show that political parties' proposals for reversals are linked to the interests of their constituencies.
Second, we argue that reversals are more frequent when transactional costs (that is the costs associated with reaching a decision) are lower. It could be argued that transactional costs vary from one type of policy to another and that structural reforms -rather than spending cuts -are most costly to revert (Lowi 1964, Wilson 1980. Finally, it must be noted that who is governing after conditionality also makes a differenceespecially considering our argument that governments have some leverage during a bail-out. When there is continuity in office, ministers are less keen to revert the reforms that they themselves passed. By contrast, if a new government is elected, it is more likely to campaign and later to undo the predecessor's reforms. Along that line, Pop-Echeles (2008) argues that, in Latin America and Eastern Europe, reversals during or after conditionality occur when a left-wing government follows a right-wing one (see also Campos and Horváth 2012). Research by Branco et al. (2019) and  also testifies to the crucial importance of partisanship turnover in explaining reversals in Spain, Portugal and Italy. Nevertheless, we claim that every government will be torn between these contradictory forces and that partial reversals will occur in all cases. 1 The risk there is one of 'inertia' -a situation in which domestic actors refuse to replace norms, rules, and practices with new ones (Radaelli 2003). Robert Schuman Centre for Advanced Studies Working Papers

Methodology
This paper and our book cover the period during and beyond the crisis (2007-2019). The case selection includes the Eurozone's five bailed out countries (Greece, Ireland, Portugal, Spain and Cyprus). Even though in Spain the conditions only concerned the financial sector, the Commission (mandated by the Council) started a programme of strict monitoring of fiscal consolidation measures and structural reforms in 2011. It explicitly linked its recommendations for change to the disbursement of the loan instalments. Spain's bail-out, and in particular its negotiation and implementation, is therefore more similar to the other Eurozone bail-outs than might appear at first glance.
To demonstrate our arguments, we rely on the congruence method to explore the constraints of conditionality, the use of these constraints by different governments, and the motivation and willingness to revert what was done during the programme (George and Bennett 2005: 200). Regarding our empirical sources, validity was ensured by triangulating different types of data. First, we drew on the primary and secondary sources that reported the policy-making processes in the period under study (academic research, press reports, investigation books by journalists, memoirs). Second, we relied on a qualitative analysis of official documents from the government, the IMF, the European Commission and trade unions. Third, we conducted face-to-face interviews (between 2014 and 2018) with major actors in policy-making during the period. We contacted all ministers and secretaries of states in key sectors (finance, economy and so on), officials from the Troika, key members of trade unions and journalists. In total, we interviewed 129 individuals 2 (including 95 former ministers, Junior Ministers or government advisors, 12 Member of parliament, 8 officials from the IMF or the EC, 10 Trade Union leaders and 4 journalists) after which saturation was reached (Büthe and Jacobs, 2015: 10). The average duration of interviews was 45 minutes. They were recorded with the authorisation of the respondent and retranscribed verbatim. Anonymity was guaranteed. We analysed all the interviews manually after categorising them by country and by period. This allowed us to go in depth in our material and to obtain a full understanding of what happened behind the closed doors of international meetings.
Secondly, we developed an original, independent and detailed data set on conditionality policies. For each country, every testable policy objective included in the summaries of the IMF and the EC MoUs were coded as a 'condition' as well as subsequent monitoring reports (this means that objectives for policy outcome were not coded, as their fulfilment was not solely dependent on governmental actions). We coded every implemented policy mentioned in these documents as a 'reform' .
We then measured congruence between conditions and reforms. On the one hand, we tracked whether the conditions had been fully, partially or not implemented by the end of the bail-out. This entailed looking for an action by the government implementing it for each of the conditions. Although the reports by the EC and the IMF were very useful, we decided to consult government official websites as well as other documents, interviews, legislation and the media to triangulate our data. In line with previous investigation on pledge fulfilment (Thomson et al. 2017), the categories of fully and partially fulfilled were merged to increase inter-coder reliability.
On the other hand, we confirmed whether or not the reforms adopted during the crisis were included in the MoU. This allowed us to examine the reforms made in the period from the crisis to the signature of the MoU and to differentiate between the countries that passed many reforms before the bail-out (like Ireland) from other countries.
As for the variables affecting bargaining power (saliency, credibility, etc.), we relied on secondary literature and on the perceptions of interviewees about those variables. In a final step, we investigated whether the reforms were partially or fully reversed during or after the programme or maintained until the end of 2019. We coded the factors that can explain the presence or absence of reversals as follows: 2 30 in Portugal, 17 in Ireland, 26 in Spain, 14 in Cyprus, 34 in Greece and 9 in Brussels. In the period between January 2014 and November 2018.
 Structural: we assigned number 1 to structural reforms. These were defined as reforms that are designed to boost productivity and improve the supply-side functioning of the economy in the medium and long term; they include labour code reforms, reduction of business and trade union monopolies, decreased regulation and red-tape, privatisation, removal of licensing, improvements to the functioning of the court, public administration reforms, enforcement of property rights etc. We assigned 0 to reforms aimed primarily at cutting spending in the short or medium term.
 Concentrated benefits: we assigned 1 when the benefits of the (possible) reversal were concentrated in a specific group of the population and/or specific actors and costs were diffused.
 Concentrated costs: we assigned 1 when the benefits of the (possible) reversal were concentrated in a specific group of the population and/or specific actors and costs were diffused.
Inter-reliability tests for this coding ranged between 0.85 and 1.

Bail outs are not a 'diktat'
In all our cases, governments tried to avoid further encroachment on their sovereignty tied to formal conditionality and passed packages of structural reforms and spending cuts to regain investors' trust.
The ECB emerged as an influential actor in this period. The fact that banks in the Eurozone relied on the ECB for liquidity gave the latter the power to pressure national governments. For example, when the Parliament rejected a one-off 'deposit levy' on all deposits in Cyprus, the ECB announced the end of emergency lending to domestic banks. The deposit levy was eventually approved. Using the same strategy, the ECB exerted heavy pressure on Irish authorities to ask for a bail-out, which they eventually did.
When the governments of Greece, Ireland, Portugal, Spain and Cyprus perceived interest rates to be unsustainable, they eventually decided to ask the EU and IMF for credit. Doing so, they came under heavy pressure as their European counterparts and the ECB, which feared contagion, insisted that they asked for financial help. In some countries bail-in was discussed (Ireland) and/or pursued (Cyprus). In all cases, the decision to ask for external help was often taken by a few ministers, and was rarely consensual domestically.
After requesting a bail-out, these countries had to negotiate and implement a series of specific reforms for each tranche of the loan. The negotiations of the conditions associated with the loan were generally hasty, lasting no more than a couple of weeks. All countries were constrained to accept policies they disliked. For example, the centre-left government in Portugal did not believe that internal devaluation was necessary for growth but it was forced to comply. In Greece, various governments considered that spending cuts were going too far but implemented many of them nevertheless, and Syriza was unable to introduce the U-turn in policies it had campaigned for during the elections and the referendum. In Cyprus, although privatisations were resented by most of the parties, the Ports Authority was in fact privatised in the end. In Ireland, the government wanted unsecured senior bondholders in two of its banks to foot the bill in 2011, but this measure was not accepted.
However, despite these constraints, governments were far from powerless: in all our cases, governments of bailed out countries were able to advocate, resist, shape or roll back some of the policies demanded by external actors. This influence was not only exerted while the MoUs were being negotiated, but also during their implementation.
Informants from all sides indeed explained to us that the negotiators from the Troika were open to the solutions proposed by governments, as long as they reached general objectives (for example, the reforms of the health and education systems in Spain and of local government administration in Greece were in fact drafted by these governments). Governments felt they could take the initiative to insert specific policies in the MoU or its revisions (for example, the Guaranteed Minimum Income or the creation of a national health system in Cyprus  (Moury and al. 2021, chapter Five).
Interviewees from all countries further reported instances of policies that the lenders wanted but which were eventually dropped from the Memoranda. In Portugal, for instance, the negotiators managed to take the reduction in the minimum wage off the table; in Ireland the corporate tax was not targeted, despite the Troika's wish to see it reformed. This was also the case of the subsidised agricultural fuel (green diesel) in Ireland, which the Troika wanted to abolish and the government managed to exclude from the programme. As an Irish high official explained to us: "the Irish side had many 'red lines' which demarcated things that we wouldn't agree on -(...) these were not public and were not crossed' (Moury et al. 2021, chapter Four).
In a similar vein, we show that domestic actors sometimes successfully resisted the implementation of some MoU policies they disliked; and that lenders were sometimes willing to turn a blind eye and let non-implementation go unsanctioned. This often happened towards the end of the programme when the economic conditions improved and/or when trust was high or even when the Troika was suffering from 'reform fatigue'. In particular, in Greece, an official from the Troika recalls about the third bail out: "At this moment (...) the development plan has ended up a mix of various things that the government wants to reverse and for which the Institutions and especially the Commission has given up and says; 'ok, go ahead', especially as far as the labour market is concerned. (Moury and al. 2021, chapter 7).
Similarly, in Spain, the pension sustainability factor was approved but never implemented. In Cyprus and Ireland, a privatisation programme included in the MoU was never fully executed and, similarly, many of the privatisations in Greece are still pending. Lastly, the Portuguese government resisted pressure from lenders to find an alternative to the Constitutional Court's annulment of a permanent reform of the pension system.

Bailouts were seen by many decision-makers as an opportunity to pass difficult reforms
Given this leeway, domestic negotiators told us that, in some cases, the bail-out represented an opportunity to pass reforms that they considered necessary but could not previously have passed. Our book demonstrates that the decision to ask for a loan, the negotiation of the MoU and its implementation were indeed centralised within the executive (and within it, in the Ministry of Finance). Only a small number of people were fully informed of the Troika's red lines, and the government in some instances exploited this asymmetry of information when presenting its plan to the parliamentary opposition or stakeholders. In many instances, this context allowed ministers to adopt reforms that they deemed necessary but could not previously have passed.
For example, for Greece, many of the issues about structural reforms raised by the Troika in the first bail out were already in PASOK's electoral programme and the Greek government saw the MoU as an opportunity to ensure their implementation (Ladi, 2014;PASOK, 2009). In his memoirs, Papakonstantinou (2016: 173) sets out a list of measures he considered positive, including the opening up of regulated professions, improvement of the business environment, liberalisation of public enterprises, public administration reforms and the reduction of the state's role in the economy generally.
Another key Greek Minister in the negotiations described MoU measures like the public administration reform and the Kallikrates local government reform as "self-evident and not imposed by the Troika".
Similarly, ministers from Portugal told us that the bail-out made the reduction of severance payment possible, as well as the decrease in the rents (excessive profits) enjoyed by the pharmaceutical and Electricity Companies; and that they had been wanting to do this for a long time. In Spain, informants from both centre-left and centre-right governments told us that they welcomed the reform of the pension system. Similarly, in Cyprus, our interviewees from all major parties stated that the social benefits system needed rationalisation. More generally, a government Minister from the Anastasiades government stated that the MoU "could thus be considered as part of and indeed as a centrepiece of our governmental programme" (Moury et al. 2021, chapter 7).
In Ireland, respondents acknowledged that the liberalisation of legal services had been on their agenda for some time. A high state official, for example, noted that "any good finance ministry will look at what are all the things that we would like to do that we haven't been able to do for a number of years and there's a crisis moment in which you can do them, for example taxation on property". Another senior civil servant told us that it was not the Troika but the government that had requested the reduction in the minimum wage (Moury and al. 2021, chapter Four).
As a result of this leverage, elections during a bail-out changed the terms of conditionality, albeit to a limited extent. In Portugal, elections took place immediately after the Socialist party had negotiated the MoU and, during the programme, new measures -such as a new section on deregulation and the further liberalisation of the labour code -were added in the reviews at the request of the centre-right government. In Ireland, the change in government formed just after the bail-out negotiations led to a reversal of the minimum wage cut and an extension of the fiscal consolidation period. In Greece, the SYRIZA government did not manage to change the conditionality of the programme but attempted to implement a 'Parallel programme' focusing on alleviating some of the social inequalities arising from austerity.
This finding consolidates the strand of literature which shows the relevance of national executives during moments of external conditionality imposed by the EU (Dyson and Featherstone 1996;Pochet 1999;Featherstone 2003 . This working paper thus also nuances the argument that the conditions of international loans are a 'diktat' (Hill 2003;Keaney 2001); or that democracy 'ended' during the crisis (Pianta 2013;Streeck 2014). We show that conditionality narrows, but does not annihilate, political choice, and that voting still matters. Even if strongly constrained, governments in all cases had alternatives to balance the accounts and room for manoeuvre as regards the structural reforms.
However, this was true only as long as there was a minimum of trust between the lenders and the bailed out government (see below). We also show that governments rarely acknowledged their leverage in public or in their discourse with stakeholders. This led to a decline in the transparency and openness of policy making in light of crisis management. Yet these are fundamental dimensions of the quality of established democracies (Altman & Pérez Liñán 2002). The reforms implemented during the crisis were highly salient and should be at the centre of public debate and choice (Seville 2017).

But some countries had more leverage than others
Our case studies in the book also reveal that in many instances executives agreed with the lenders on the diagnostics and solutions -a phenomenon that was underlined by the fact that technocrats or 'sympathetic ministers' were often appointed during the bail-out (as shown by Alexiadou and Gunaydin 2019). When there was disagreement, the leverage of the bailed out executives was, according to interviewees, dependent on five inter-linked variables: economic fundamentals, the costs for lenders of not granting assistance, domestic governments' credibility and insider knowledge, and the saliency of the issues at stake. Robert Schuman Centre for Advanced Studies Working Papers First, the evolution of the economic conditions and the urgency of striking a deal (Offe 2017;Finke and Bailer 2019;Schimmelfennig 2015;Hennessy 2017) were important determinants of a country's leverage. Indeed, informants from bailed out countries and from the Troika reported to us that a change in the economic fundamentals of the bailed out countries (principally debt and yields) impacted their bargaining power. When the economic situation of any country deteriorated so much that it had no choice but to ask for a bail-out, its bargaining position inevitably weakened. When domestic banks were in trouble, then governments became very vulnerable to the ECB's cut of liquidity (as happened in Cyprus and Greece) or to the threat of doing so (as in Ireland). Alternatively, if the economy of bailed out countries improved during the programme, they became less reliant on the loan, and their bargaining power was stronger. Towards the end of the programme in Portugal, for example, (when the economy was recovering and the yields were at their lowest point for a decade), the government was able to say no to the lenders' requests to introduce an alternative to the constitutional court's annulment of a permanent reform to the pension system. Similarly, in Cyprus, the last review was never concluded since the economy was already recovering and Cyprus decided not to absorb all the available funds.
Second, the comparison between cases shows that the bargaining power of domestic executives is positively related to the costs lenders would pay if they failed to grant assistance (Howarth and Quaglia 2016;Hennessy 2017). Countries like Spain that were 'too big to fail' were in a better position than smaller ones such as Cyprus. By the same token, Greece was in a stronger position in 2012, when the cost of a default would be high for German and French banks and the Eurozone as a whole, than in 2015 when the Eurozone had already started reforming and it was much more protected from a potential Grexit.
Third, our analysis also stresses that the bailed out governments' credibility is important to leverage, a variable which has been understudied (but see Sacchi 2018). Building credibility during the crisis was a complex, multi-faceted issue. The improvement or deterioration of economic fundamentals was one factor, but our research reveals that the perception of these changes was also paramount. A good example of this is the successful strategy pursued by the Irish, Portuguese and Cypriot ministers to 'under-promise and over-deliver', that is to purposefully keep expectations low as regards the reduction of their deficit so that they could then do better than expected. As they explained to us, the 'surprise effect' had a positive impact on the EU actors' (and investors') evaluation of a given country's efforts, and hence in turn increased their credibility.
Credibility also derived from a given government's capacity to implement the agreed deals, itself facilitated by political cohesion and stability. Credibility also interlinks with personal trust, that is the good faith and competence of the interlocutor as perceived by the lenders. This was particularly visible in Greece, where the credibility of the government (and hence its leverage) varied at different times depending on the success and speed of the implementation of reforms; and where the 'unorthodox' behaviour of ministers such as Yannis Varoufakis quickly created distrust among their international partners and annihilated Greek leverage.
By contrast, the lenders perceived ministers and top civil servants in Ireland and Portugal as being qualified and motivated to implement the deals, and this gave extra leverage to the domestic negotiators. Similarly, in Spain, the absolute majority of the PP, which guaranteed the approval of all reforms without parliamentary vetoes, and the government's commitment to external demands increased their credibility and room for manoeuvre vis-à-vis the lenders. Cypriot negotiators were seen at first with suspicion because of their links with Russia; but later were considered more credible given the national consensus around the reforms.
Credibility and trust, moreover, evolves with time. Notably, the Portuguese Minister of Economy notes in his memoir that: "It is important to note that even the position of the Troika changed over the years. At first, the Troika was extremely demanding and intransigent on certain matters; indeed, there had been a lot of mistrust on the part of international creditors, (...) However, at a certain point the Troika started to be much more flexible and very understanding". (Santos Pereira 2014:103). In this context, credibility and partisanship are related insofar as the congruence of the international lenders' and domestic ministers' ideology (and that of top public servants) increases the trustworthiness and consequently the latter's leeway. In Cyprus, Portugal, Spain and Ireland, particularly, the lenders and the debtors were generally in strong agreement on the country's problems and how to resolve them. This gave domestic negotiators leeway not only to determine the policies to be followed, but also not to implement some of the measures that had been agreed. The situation was the exact opposite at the start of Syriza's governance (2015) when trust between the institutions and the Greek government was at its lowest 3 . However, the relationship between credibility and partisanship is not straightforward: the Portuguese centre-left government was highly trusted after the bail-out due to its ability both to keep their radical left partners on board and respect the EU rule. This also happened in SYRIZA's second term of office in Greece, when the government started implementing the reforms without creating strong social unrest.
Credibility seems to increase when experts and technocrats occupy important governing positions. This was definitely the case in Greece during the Papademos interim government when the negotiations with the Troika are described by our interviewees as easier since Papademos and the lenders spoke the same language and shared the same network. Cyprus and Portugal tried to increase their leverage by appointing technocrats as Ministers of Finance. These results serve as a reminder for governments that, in future crises, building credibility is of paramount importance to strengthening one's hand in negotiation.
Insider knowledge is a fourth variable that was crucial to increase the government's leverage in negotiations with the lenders. While the lenders' representatives had specific ideas on policies to be implemented by debtor countries, they lacked information about the domestic context. This opened up the possibility for national governments to influence the design of solutions to the crisis. Nevertheless, we show in our book that insider knowledge was more valued by the lenders when trust between the two parties was high.
Finally, the saliency of the measures was also crucial in determining national governments' room for manoeuvre. Whenever a policy was considered to be a priority by the lenders, the executives' capacity to influence the final outcome was much more limited. All the labour market reform measures were central for the IMF, whereas the liberalisation of services was the priority for the EC; but both lenders insisted on cutting spending to allow for immediate macroeconomic adjustment. Although governments were never powerless, in all those cases they had much less room for manoeuvre and executives were much more likely to be forced to accept measures against their will. By contrast, governments felt much freer to shape the policies for many structural reforms (for instance in the field of justice, health and education, and in some cases banking reforms).

MoU's are less similar than would you think
Another important observation made in the book is that MoUs for the different countries vary. While all documents are based on the same assumptions (the need to cut spending, privatise, introduce structural reforms, deleverage banks and internal devaluation), the solutions proposed, the details of the measures and the stringency of conditions all differ. This result then nuances the claim that IOs use a one size fits all model (Rodrik, Easterly and Krueger 1999).
This variety is a direct consequence of the fact that governments had leverage to identify the structural problems the country was facing and to propose solutions addressing those problems. Robert Schuman Centre for Advanced Studies Working Papers We also observed differences in the precision and the number of conditions. Different explanations for longer MoUs have been advanced in the literature, such as the economic weakness of a country prior to the crisis (Copelovitch 2010;Dreher 2004). This appears to be true in our cases, but our processtracing also uncovers other explanations that had not yet been explored. Again, trust seems to be of paramount importance, insofar as when lenders do not believe the decision-makers in a given borrower country act in good faith they will tend to detail the policies in writing (and hence insert more conditions). Another variable suggested by our process-tracing analysis is the fact that some measures were set out in detail in the original MoU or its review at the request of the government itself, to more easily circumvent the opposition to those reforms. In Spain, for example, the interviewees from the governments mentioned that they included the liberalisation of some professions in the MoU hoping to bypass veto players. In Portugal, some parts of the labour code were specified in great detail in one of the reviews at the request of the government and for the same reason. This result is in line with Vreeland's (2005) finding that the more the veto players, the more the conditions in the MoU.

MoUs are implemented to a large extent
Given that MoUs included policies wanted by the government as well as others imposed by the lenders, to what extent were they implemented? Our research shows that, despite the struggles between national executives and the Troika, a large number of MoU reforms had been (at least partially) implemented. Indeed, the spending cuts and structural reforms made by the five countries in our analysis were mostly in line with the MoUs. These findings are congruent with studies that have shown that IMF intervention is efficient when it comes to decreasing public spending on health and education (Nooruddin & Simmons 2006).
In Table 1, we present the number (and percentage) of important conditions included in MoUs, whether or not they were (at least partially) implemented and whether or not measures had been reverted during the programme (see next section for post-programme reversals). In Table 1, we see first that measures deriving from the MoUs were occasionally reverted during the bail-out. This was the case seven times in Greece, and once in Ireland. For example, in Greece, the Hellenic National Broadcasting Corporation re-opened when SYRIZA took over. In Ireland, the 2010 minimum wage cut agreed with the Troika was reverted in 2011 by a new government. This, again, testifies to the fact that domestic governments had some degree of room for manoeuvre during the programme.
On the other hand, Table 1 also shows that non-implementation is relatively rare: 84% on average and at least 75% of the conditions of the MoUs were at least partially implemented during the bail-outs 4 . This is in part because governments 'own' some of the reforms included in the MoUs, but also because not implementing the MoU is costly. We do, however, observe variation in the fulfilment rate: Ireland and Spain are the champions of fulfilment (more than 90%), followed by Greece (86%), Cyprus (85%) and Portugal (75%). The Spanish and Irish rate is easily explained by the fact that there was a great congruence of views between the centre-right governments and the lenders. Moreover, Ireland was already very liberalised. In addition, the Irish MoU contains less specific policy statements than its counterparts (19 versus 75 for Portugal, for example); it focuses more on objectives to be reached than specific measures. And, as we have seen in Cyprus, there was a great internal consensus about the need to introduce the reforms.
The results for Greece and Portugal are more surprising. In Greece, although the high fulfilment rate goes against conventional wisdom, it is linked to the fact that the breadth and the duration of the programmes were much greater than in the other countries and, therefore, some reforms started in the first MoU and were completed during the second and third one. Additionally, when conditions were not 4 Important statements are those present in any of the EC or IMF summaries of the MoU reviews (and European Semesters in Spain). Only actions are coded (not outcomes). Full and partial fulfilment were merged due to low inter-coder interreliability to distinguish between full and partial fulfilment, and in line with choices made by pledges fulfilment scholars. met immediately, they were broken down into smaller measures so that their fulfilment could be more easily monitored. Finally, the Greek case shows that the country's disagreements with the Troika were very often about the pace and depth of the reforms rather than the substance (e.g. there was agreement on the need for the pension system reform but the Greek governments could not fulfil the reform at the fast pace required by the Troika).
Taking a closer look at the Greek data, we calculated the difference between the implementation rates of each MoU. Whereas the fulfilment of the first MoU was high (39/48, or 81 % of the conditions were at least partially fulfilled at the end of the first programme), the scores are lower at the end of second (41/66 or 62%) and very high for the third MoU (59/64, 92%).
Finally, Portugal had the lowest fulfilment score, despite the fact that ministers generally agreed with the lenders about the country's diagnostics and prognosis (see chapter five of the book). One obvious reason for this lower fulfillment is the fact that the Constitutional Court blocked the adoption of 9 measures in the MoU; if this had not been the case, Portugal's implementation rate would have been more similar to the other countries. In relation to this, our five cases also show that conditions were often only partially met because of the strength of special interests -see for example the liberalisation of professions in Greece, Ireland and Spain; or the decrease in rents in the energy sector in Portugal. Those findings are congruent with the research of Juwana (2005) who stresses the roles of Courts, and of Ivanova et al (2003) who state that political systems with strong vested interests were less likely to deliver on externally-mandated reform programmes (see also Stolfi et. al. 2019 for Greece). Due to our focus on governmental manoeuvring, a limitation of our database, is that it addresses only reforms or spending cuts. Deficit reduction objectives were not included. Hence, our quantitative findings do not necessarily imply that the conditions fully reached their objectives. What our data shows, however, is that the adoption of MoUs and adjustment programmes -which are to some extent 'owned' by governments -proved to be an effective model for ensuring the adoption of specific policies. Robert Schuman Centre for Advanced Studies Working Papers

Reforms adopted during conditionality are quite resilient
A much neglected question in the conditionality literature is the resilience of reforms after conditionality. In Table 2, we present the absolute number and percentage of the measures passed after the crisis that were reverted once the programmes ended (to December 2019). Therefore, Table 2 not only shows the reforms included in the MoUs but also those passed by governments just before the bailout, under pressure from external actors.
We can see that reversals occurred in all countries, as we theoretically expected. However, the extent of reversals varies across countries. In particular, post-programme reversals are rare 5 in Greece and Cyprus. This is not surprising in Greece given that it terminated the third bail-out in August 2018, just 16 months before the end of our analysis. Furthermore, post-conditionality scrutiny has been stricter in Greece than in the other four countries because the Greek loan was the biggest of all and there were many more reforms to be followed up by the government in order to guarantee a long-lasting economic recovery. However, some reversals had been planned in the 2020 state budget prepared by the Syrizaled coalition government (and subsequently implemented by the Conservative government led by New Democracy). This is the case, for example, of the reductions in selected VAT rates, the re-introduction of the 13th month payment for pensioners, and a reversal of an earlier reform of survivors' pensions. In Cyprus, the rarity of reversals could also be explained by the fact that the bail-out terminated later than in Portugal, Spain and Ireland, but also by political continuity -as Nicos Anastasiades has been the Cypriot President since early 2013. Nevertheless, some reversals were also made there -such as a progressive reversal of the cuts in public servants wages.
In Ireland, Spain, and above all Portugal, post-programme reversals were more frequent. In Portugal, in particular, 50% of the measures taken under conditionality were later dismantled, mostly by the minority Socialist government that was supported by three radical left parties that had put reversals as a condition for their support De Giorgi and Cancela 2019). In Spain, the new minority Socialist government had promised during its campaign to pass many reversals, but political instability prevented it from passing many of them. In Ireland, there were many reversals despite relative political stability (Fine Gael governed both during and after the bail-out, first in coalition and then alone): most of the public sector cuts and some social benefits cuts were, at least, partially reverted. In this country, many reforms had been taken before the bail-out, and hence most of the measures adopted during the crisis were related to fiscal consolidation, which means that they were also easier to revert in the post-programme period. Table 2 also shows that only a minority of reversals were passed by governments of a different persuasion. This means that reversals will occur even if there is political continuity. Table 2 further distinguishes structural reforms and other types of measures and shows that out of the 58 reversals, only 10 were for structural reforms. This number is mainly driven by the Portuguese case (7 out of 10), where many parts of the labour code reforms were eventually reverted (although the central parts remain, see . From our data, we also calculated that the great majority (91%) of structural reforms enacted during the crisis remained. For example, the labour market reforms in Spain, the reforms of the banking system in Cyprus, the introduction of the Public Revenue Independent Authority in Greece and the public administration reform or the pension age increase in Ireland were all kept. These figures and the information in the case studies testify to the governments' unwillingness to revert structural reforms.
Finally, in Table 3 we look at the correlation between the type of policy and the likelihood of reversals. Here, we see that reversals are less likely when the reforms were structural and involved concentrated costs, and more likely when they concentrated benefits and diffused costs. These differences are statistically significant. The findings speak to the conclusion reached in the literature that governments make cost-benefit calculations when deciding which measures to revert (  Our analysis thus shows that since the end of Eurozone bail-out programmes, executives have been able to revert what was done during the programme. However, these reversals were partial and limited. Reversals mostly concerned visible spending cuts and tax increases, while less noticeable changes in taxation were often kept. We also show that the great majority of structural reforms were maintained, and thus that conditionality has at least a medium-term effect. The bail-out has therefore brought about a profound and durable change in the socio-economic structure of the society, even though the national growth models persist (Chang, Steinberg and Torres 2019). For example, the labour code is very different today than it was a decade ago in many countries (see also Afonso 2019; Moury and Afonso 2019;and Branco et al. 2019). Pensions have seen profound reforms in some countries and durable spending cuts were made (in the health sector for example). The leverage ratio (that is the percentage of debt in the banks' balance sheet) has decreased significantly. Many professions have been liberalised, rents in the pharmacy and energy sectors have been (at least partially) cut and various businesses and services have been privatised. Further research should certainly cover the resilience of reforms in a longer time lapse. These results again show that conditionality can be effective; but also that spending cuts and tax increases are less resilient than structural reforms. This difference is again the product of the varying degrees of ownership by domestic governments which was larger for structural reforms. Not surprisingly, this suggests that reforms are more durable when they correspond to the government's preferences.

Conclusions
How did bail-out countries deal with the challenges of the Eurozone crisis over those 13 years? The main argument of our book and of this working paper is that domestic actors are also capable of extracting policy benefits from a bail-out both during and after the adjustment programme. We start by showing that, contrary to what it might appear, a bail-out is not always a diktat, it can also be a twosided process in which executives negotiate with international lenders. Bailed out governments keep some leverage, we show, because the lenders value 'ownership' and 'insider knowledge'. Indeed, the IMF recognises that governments often do not implement unilaterally imposed deals. Moreover, lenders are generally not experts of a particular country so need insider knowledge to understand the causes of the crisis and design efficient solutions.
Considering this domestic leverage, we show that elections make a difference during a bail-out, as governments of different political persuasion are sometimes able to renegotiate or resist the implementation of the deals agreed by their predecessors. We further illustrate that conditionality is to some extent useful to executives. This is because the negotiations of the bail-out take place behind closed doors and the costs of its rejection are very high. This sometimes allows executives to pass reforms that they deem necessary but could not have passed in the past because they were very unpopular or were blocked by powerful interest groups.
The leverage of governments -and their capacity to pass reforms corresponding to their preferences -inevitably varies from case to case and is dependent on the bargaining context. We show that the degree of external credibility of domestic actors is a crucial variable and, in itself, a variable of the willingness and capacity of the government to implement the agreed deals. When the lenders think that the government shares their main assumptions and is in a position to fulfill what has been agreed, its leverage is higher.
We finally demonstrate that, after a bail-out, executives are both able and willing to revert what was done under conditionality. They are able to do so because the lenders do not have the same 'sticks' and 'carrots' as during the programme (Drazen 2002, Vreeland 2003. Given this ability, governments selectively pick reversals that bring them more electoral benefits at a lower cost.
Our findings about the potential but also the limits of conditionality are important for explaining domestic variation during the Eurozone crisis but also informing policy-makers of what works and what doesn't work when tackling the new European economic crisis linked with the Covid-19 pandemic.