Estudio de crecimiento anual independiente. 5.º informe EPACi 2017 Fit for purpose: a German-Spanish proposal for a robust European Unemployment Insurance Sebastian Dullien, Jonas Fernández, Marc López, Gero Maass, Daniel del Prado, Jakob von Weizsäcker Contents Ÿ Ÿ .............................................................................................................................................................. . 2 Ÿ Ÿ A European Unemployment Insurance fit for purpose............................................................. 3 Ÿ Ÿ  to stabilize the euro area economy without creating political discord: a compromise proposal for a European Unemployment Insurance Scheme ................ 5 www.fes-madrid.org 1

Preface How do we contribute to fixing the Euro and fighting the North-South divide in growth and social welfare? And how do we change the debate on the future of the euro area from one fought on partisan stability versus transfer lines, to one seeking a common solution, which both safeguards the long-term viability of the European project, while also taking into ac count individual Member State interests? These questions have been the driving force behind a German-Spanish working group, ini tiated by Agenda Pública and the Friedrich Ebert Stiftung. Over the course of several meet ings, the present report by Professor Sebastian Dullien and Professor Daniel del Prado on a European unemployment insurance scheme was developed, taking into account both the principles of solidarity and of individual responsibility. We thank the Members of the European Parliament Jonas Fernández and Jakob von Weizsäcker for backing and developing the idea, and, crucially, for lending their political support. We look forward to the coming debates! Gero Maass Head to the FES office in Madrid Marc López Director of Agenda Pública 2 www.fes-madrid.org

A European Unemployment Insurance fit for purpose Jakob von Weizsäcker(Member of the European Parliament) and Jonás Fernández(Member of the European Parliament) Fixing the euro remains one of Europe’s most pressing policy challenges. The broadening cycli cal recovery in the euro area is very much welcome but must not lead to complacency in that respect. Strengthening both responsibility and solidarity at the same time will be essential for creating a sustainable euro area architecture. Obviously, this implies that the euro area is not at its possibility frontier, where more responsibility would mean less solidarity or vice versa. Of course, it is possible to have a currency union without solidarity. But that would be akin to building a car without a suspension and with seats made out of steel. It would be possible but very uncomfortable since all the shock absorption would have to be done by the passengers’ spines. Many citizens in the euro area member states undergoing large economic shocks must have felt that the euro area is akin to that sort of misconstrued car. They would not be entirely mistaken, since one key shock absorber is absent by definition in a currency union, namely the exchange rate. And a second shock absorber, fiscal policy, did not work well in many euro area countries where public or private debt levels were already very high to start with. One possible remedy would be to turn the euro area into a full-scale fiscal union. However, a number of member states are adamant that they do not wish to go down that route out of fear that this could result in large-scale permanent transfers. Therefore, solidarity by means of large-scale transfers is not a convincing solution for the future of the euro area. Instead, the euro area will require solidarity that is based on a fair system of insurance where no Member State will have to feel unduly disadvantaged in the long run. And politically, by far the most attractive form of“solidarity” insurance is unemployment insurance since it tackles head on the worst and most visible consequence of large scale economic shocks and alleviates the fiscal strain in bad times in the context of the Stability and Growth Pact. However, designing a full-blown and fair system for euro area unemployment insurance would require a substantial harmonisation of euro area member states’ labour market regulation and welfare systems which is definitely not on the cards for the time being. Therefore, a slightly less ambitious model that still offers sufficient shock absorption ca pacity should be envisaged. Designing such a system, that works both politically and eco nomically, is precisely what we have been working on in close cooperation with Sebastian Dullien and Daniel Pérez del Prado. www.fes-madrid.org 3

A European Unemployment Insurance fit for purpose From insurance theory we know that ordinary insurance across a longer period of time can typically be closely modelled as a mix of self-insurance and reinsurance. For exam ple, in car insurance a significant part of the insurance is self-insurance. In fact, after an accident, the insurance premium increases, allowing for a substantial share of the loss es to be paid back to the insurance company over time. On the other hand, in case of accidents with very high damage, the losses incurred are absorbed to a large extent by the insurance community. This basic insight from insurance theory inspired our proposal for a workable euro area unemployment insurance framework. In normal times, euro area member states would pay 0.1% of GDP per year into a common European unemployment fund. The lion share of this would go into a national compartment earmarked specifically for this country which is the self-insurance compartment. The rest would go into a common“stormy day” compartment for very large shocks for the purpose of re-insurance. If a Member State experiences a rise in unemployment over a set reference value(say 0.2 percentage points), it would receive a net pay-out from its national compartment to support the increased unemployment benefits. If a country is hit by a very large economic shock(say over 2 percentage point rise in unemployment), it would receive additional pay ments from the stormy day fund as re-insurance. By excluding net-payments into the system from the Stability and Growth Pact in good times, the fiscal restriction in good times would de facto become tighter. Conversely, in bad times the net-payments out of the system would also not be counted for the purpose of the Stability and Growth Pact so that these extra funds would relax the overall fiscal constraint of a country in shock. In this way, the system would contribute significantly to more reliable and credible fiscal stabilisation, in the face of asymmetric and even to some extent symmetric shocks. The extent to which the different compartments would be allowed to run a deficit in order to enhance the stabilisation effect beyond the workings of a pure rainy-day fund will de pend on the credibility of the overall institutional set-up. In any event, member states with deficits in their national compartments would be required to make higher contributions once their economies recover. Simulations based on these principles show a significant economic stabilisation potential from the system, with minimal net costs to its contributors over time. This form of unem ployment insurance would be as much an institutionalisation of counter-cyclical economic policy, as it would be a form of solidarity that would assure a much smoother ride for the citizens of the euro area. 4 www.fes-madrid.org

How to stabilize the euro area economy without creating political discord: a compromise proposal for a European Unemployment Insurance Scheme Sebastian Dullien(Professor of International Economics at HTW Berlin, the University of Applied Sciences) and Daniel Pérez del Prado(Professor of International Social and Private Law at Universidad Carlos III of Madrid) Introduction By now, the need for a fiscal capacity of the euro area is widely acknowledged. 1 While the original euro area framework in the Maastricht Treaty was designed without a fis cal capacity based on the belief that, if countries had solid public finances, they could self-insure against shocks by borrowing in international markets in a crisis and using the funds for national stabilization policies, the consensus today is that this hope was overly optimistic. First, many euro member states did not run a counter-cyclical policy as one would have needed prior to the global financial and economic crisis of 2008/9 and hence saw themselves cut off financial markets. In some cases(such as Greece), the failure to do so was in clear violation of the EU’s fiscal rules. In other cases(such as Spain or Ire land), the failure was mostly based on incorrect judgements of the economies’ position in the cycle. Second, in the case of banking crises, asymmetric shocks in some countries (notably Spain and Ireland) proved so large that these countries lost(or were at the verge of losing) access to financial markets even though they had entered the crisis with very sound public finances. A fiscal capacity is now seen as a tool to first help countries stabilize their economies when hit with an asymmetric shock or with a symmetric shock monetary policy is unable to counteract, and second to make fiscal policy overall more counter-cyclical. 1 As evidence, see the Four Presidents’ report of 2012, the Five Presidents’ report of 2015, the proposals by the new French president Emmanuel Macron as well as the Spanish government from 2017. www.fes-madrid.org 5

How to stabilize the euro area economy without creating political discord... One of the elements which could form part of such a fiscal capacity is a European Unem ployment Insurance(EUI) or synonymously a European Unemployment Benefit Scheme (EUBS), for which by now a number of variants have been spelled out and about which a number of studies have been conducted. 2 Such a EUBS would result in a system that stabilizes aggregate demand in a country in which unemployment increases strongly in the wake of an economic downturn, either by directly paying funds to the unemployed from a European fund(under a so-called genuine scheme) 3 or by making transfers to national unemployment insurances(under a so-called equivalent or reinsurance scheme). 4 Yet, there are a number of concerns about such a system: especially the German estab lishment is afraid that introducing such a system could lead to permanent transfers from Germany(and other“Northern” euro-countries) to the euro-periphery and countries such as Italy, Spain, Portugal, France and Greece. Another concern is moral hazard: Governments might not do enough to fight unemployment as costs are shared(Feld and Osterloh, 2013; Wissenschaftlicher Beirat beim Bundesministerium der Finanzen, 2016). Moreover, some variants of a EUBS might require treaty change and are hence seen as toxic in some coun tries. Finally, social partners in some countries fear that the introduction of a European unemployment insurance might weaken the tripartite governance structure of successful national unemployment benefit systems. Against the background of this debate, this paper is spelling out a compromise proposal which on the one hand would provide the stabilization sought by proponents of a EUBS, but on the other hand addresses some of the most important concerns brought forward against a EUBS. Before going into the details of such a proposal, it is necessary to define the economic aim of such a scheme. The aim of the proposal discussed here is two-fold: It should first lead to cyclical stabilization both at a country-level as well as an EMU-wide level by influenc ing the pattern of national aggregate demand over time. Second, in the case of very large shocks to individual countries, it is supposed to soften the impact of this shock. What the proposal is not supposed(or able) to do, however, is to provide long-term convergence of incomes in different euro area countries. 2 The most comprehensive study of which has been conducted for the European Commission by Beblavý and Lenaerts(2017). 3 A comprehensive proposal for such a system can be found in Dullien(2014). 4 The standard reference for such a proposal is Beblavý et al.(2015). 6 www.fes-madrid.org

Sebastian Dullien and Daniel Pérez del Prado Our proposal borrows elements from a number of published proposals, among oth ers those from the CEPS(Beblavý and Lenaerts, 2017) and the Italian treasury(Ministero dell’Economia e delle Finanze, August 2016, September 2016a, September 2016b). We be lieve, however, that the combination of the specific details is new and maximise the stabi lization ability while minimizing concerns among euro area member states. Especially, our system has significant elements of self-insurance, by which countries are forced to make savings in good times for the case of a severe down-turn, combined with elements of joint insurance(or solidarity), by which countries especially hard hit by a crisis get support from a commonly financed fund. The rest of the paper is structured as follows: section 2 presents the basic elements of the compromise proposal, section 3 presents some financial data from a simulation exercise of the proposal and section 4 covers legal questions on the implementation of such a system. Basic elements of this compromise scheme While there are some advantages of designing a European unemployment insurance as a genuine system under individual workers acquire claims against a European unemploy ment insurance, the technical and legal complications of introducing such a system are immense(see for the legal issues below). Therefore, a compromise proposal should be de signed as an equivalent system under which payments are made out of a country’s budget and transfers are sent to the country’s budget. Each country would then be responsible to organize the transfer of the funds between its general budget and its unemployment insurance system. Under such a scheme, each country would pay in 0.1 percent of its GDP each year into a common European unemployment fund or budget line. 80 percent of the country’s pay-ins would be earmarked in a national compartment, the other 20 percent would go into a com mon compartment for very large shocks(a“stormy day fund”). If a country’s compartment has accumulated 1 percent of the country’s GDP, contributions stop until the reserves fall below this threshold again. Pay-outs will be made if the unemployment rate increases by more than 0.2 percentage points relative to the average of the past five years. This rather small trigger is chosen as larger triggers bring about a number of problems, such as countries potentially not know ing for a long time whether they will receive any payments at all(and hence not being able www.fes-madrid.org 7

How to stabilize the euro area economy without creating political discord... to plan with the funds in their budget) or governments having an incentive to conduct policies which just would lift the unemployment rate above a certain trigger. 5 If unemployment increases by more than 0.2 percentage points, countries can draw mon ey from their national compartments. These pay-outs would be linked to the costs of unemployment in the country concerned, and could be set at a share of 25 percent of average wages paid per employee. In order to limit the fund’s transfers, countries can de cide in the wake of small fluctuations not to draw from the fund. Such a decision would lead to less contributions in the future(as the national compartment would more quickly be filled up again). If a country were hit by a very large shock, defined as increases in the unemployment rate of 2 percentage points or more, additional payments would be made from the“stormy day fund”, the common compartment for very large shocks. These payments would be con structed in a progressive way with transfers becoming proportionally bigger the larger the increase in unemployment. This progressive phasing in of these extraordinary transfers would prevent any moral hazard problems with fixed trigger values. Disbursements from both the national compartment and the common fund should be made quickly at the point when official estimates first project an increase above the trig ger values in order to make sure that countries can work with the funds available and the payments do not become pro-cyclical. Overpayment under this set-up would not be a large problem: If a recession turns out to be less severe than projected, this would result in an excessive drawing down of the national compartment which in turn would result in a more prolonged period of contributions in the future. Excessive payments from the stormy day fund would have to be replenished from the national compartment. Each country would be allowed to run a cumulative deficit in its national compartment of up to 5 percent of its GDP. This deficit is first financed by loans from other national com partments, then, if all funds are exhausted by borrowing of the European unemployment fund in financial markets. The system would thus be allowed to issue bonds, backed by future contributions as collateral. In order to make the use of this possibility very unlikely, countries would only be allowed to participate in the system once their national compartment has collected 1.25 percent of GDP. 6 5 For a discussion of these problems, see Dullien(2017). 6 To make sure that this amount is reached quickly, one could envision a phase-in-period in which countries pay in more than their usual national contribution. 8 www.fes-madrid.org

Sebastian Dullien and Daniel Pérez del Prado In order to counter the fear of permanent transfers, a risk-based adjustment of contri bution rates would be part of the system: Under such a rule, countries with cumulative deficits of more than 0.5 percent of their GDP in their national compartments would be required to pay higher contributions as soon as the unemployment rate falls strongly(i.e. more than 0.5 percentage points compared to the average of the past three years). The increased contributions would be designed in a progressive way, with larger contributions when unemployment drops very quickly and a formula which prevents an abrupt increase in contributions. Such a dynamic claw-back would help to limit the problems often dis cussed in the context of such rules that they run the risk of taking money from countries in long recessions and hence limiting the stabilization benefits of such a system. The combination of national compartments and a dynamic claw-back system would limit issues of moral hazard as they would leave much of the final costs of excessive unemploy ment with the member states while allowing more stabilization as in the past. Finally, as a governance structure of the system, we would propose a tripartite system, with a board with representatives of unions, employer federations and national governments and Union’s institutions overseeing the implementation of the system. Such a structure has the advantage of minimizing concerns that the system will be run against the interests of the social partners. Financial performance and stabilization impact of this scheme Running a simple simulation of the proposed system against historical data shows that the system is financially stable, provides significant stabilization for some countries es pecially hard hit by the crisis while limiting net transfers to a very small amount. 7 Figure 1 illustrates the stabilization impact in the case of Spain. Spain would have received large payments out of the system between 2008 and 2013, in some years reaching almost 2.5 per cent of GDP. Under reasonable assumptions, 8 this would have transferred into GDP staying 2.5 percent higher than it historically was – a very large stabilization impact. As Figure 2 shows, the stabilization impact for Italy would have been smaller(mostly because of the more muted increase of unemployment in Italy), but still significant. 7 For the following, it has been assumed that the system started in 1997 with all euro area countries having already filled up their national compartments to 1.25 percent. The system is simulated as if all current euromember states would have been member of the European Unemployment Insurance over the whole period of time. 8 As common in the literature, it was assumed that all the funds paid out by the system were spent by the national governments and that the multiplier for these spendings was 1. www.fes-madrid.org 9

How to stabilize the euro area economy without creating political discord... Figure 1 Net payments into EUI system, Spain, 1999-2018 Spain Figure 2 Net payments into EUI system, Italy, 1999-2018 Italy 10 www.fes-madrid.org

Sebastian Dullien and Daniel Pérez del Prado Figure 3 summarizes the total finances of the national compartments(presented together) and the stormy day fund. While the global financial crisis and the following euro crisis would have depleted the reserves almost completely, overall the system would not have had to borrow . Reserves would have reached a bit less of€150 billion euros prior to the global financial crisis of 2008 and would have reached a minimum of less than€10 billion in 2014. Figure 3 Reserves in the EUI system, 1996-2018 Overall, one can see that net transfers in the system would have been limited. Table 1 presents the overall finances of the system. While in 2018, some individual countries would have significant balances in their national compartments which would need to be filled up again in following years(such as the case of Spain) or would provide funds to draw from (such as the case of Germany or France), net transfers(without the balances in the national funds) would have been rather small. As one can see, over the whole 22 years of hypotheti cal operation, the total net transfers for Germany would have been less than 0.5 percent of 2018 GDP, which translates into a mere 0.02 percent of 2018 GDP per year of the systems existence. Interestingly, France actually would have contributed more in net terms relative to GDP, and Italy, often seen in Germany as a candidate for receiving permanent transfers, would have belonged to the group of net payers. www.fes-madrid.org 11

How to stabilize the euro area economy without creating political discord... Table 1 Summary finances EUI system National compartment 2018 in bn€ Total net transfer payments 1996 to 2018, in% of 2018 GDP Euro area(19 countries) 29.0 Belgium 5.8 Germany 47.0 Estonia 0.0 Ireland-2.0 Greece-8.4 Spain-45.0 France 23.7 Italy 1.4 Cyprus-0.5 Latvia-0.2 Lithuania-0.2 Luxembourg 0.5 Malta 0.1 Netherlands 0.9 Austria 4.7 Portugal-2.4 Slovenia-0.1 Slovakia 0.6 Finland 3.0 0.45 0.47 -0.21 -0.26 -1.69 -1.69 0.48 0.29 -0.58 -0.39 -0.29 0.32 0.32 0.42 0.45 0.08 0.25 0.35 0.46 Annual average net transfer payments, 1996 to 2018 in% of 2018 GDP 0.02 0.02 -0.01 -0.01 -0.08 -0.08 0.02 0.01 -0.03 -0.02 -0.01 0.01 0.01 0.02 0.02 0.00 0.01 0.02 0.02 Maximum annual stabilization in% of GDP 0.05 0.25 1.79 1.27 2.62 2.19 0.12 0.58 1.63 1.83 1.75 0.13 0.41 0.07 0.85 0.51 0.17 0.08 Stormy day Fund 2018 in bn€ 13.70 While for the lack of consistent data, it is difficult to simulate the system for a longer time period than going back to the mid nineties, one can at least use unemployment data to evaluate how often countries would have been beneficiaries from the“stormy day fund”(which represents the transfer or solidarity element of the system) had the system existed from the sixties. Dividing the euro area countries for which we have data from 1960 roughly into“Northern” countries(Austria, Belgium, Germany, Finland, Neth erlands, Ireland) and“Southern” countries(France, Greece, Spain, Italy, Portugal) shows 12 www.fes-madrid.org

Sebastian Dullien and Daniel Pérez del Prado that over the long run, both groups would have almost equally often drawn from this solidarity fund: the Northern countries drew 41 times from the fund while the Southern countries did so 48 times(see Figure 4 ). Figure 4 Use of stormy day fund by country group Possible legal frameworks for this compromise scheme The following discussion of the legal aspects of implementing the compromise scheme we propose in this paper contains elements that have been adapted from various prior studies. While the existing literature is limited, it does document a gradual evolution in thinking on the subject. Whereas initial analyses concluded it would not be possible to create an EUBS without amending EU Treaties(European Commission, 2012; Fuchs, 2013; Repasi, 2013) 9 more recent studies assert that most of the plans proposed to date could be anchored in existing primary and secondary EU legislation, and that no changes to the Treaties would be required. The feasibility of pursuing this option, however, depends as much on the type of scheme implemented as on the way in which it is legally structured. Given the legal obstacles to implementing a genuine scheme, an equivalent scheme would be an obviously better choice(Beblavý and Lenaerts, 2017). 9 Repasi expresses certain doubts regarding the opportunities offered by Articles 352 and 153(1) TFEU. www.fes-madrid.org 13

How to stabilize the euro area economy without creating political discord... Thus, working within the framework of the Treaties and building upon existing EU legal mechanisms would be advantageous approach as it would provide a solution while sup posing, at most, reforming secondary legislation. For the sake of clarity, we have divided our assessment of existing institutions that could possibly provide a legal framework for a European-level unemployment benefits scheme into two sections: one focusing on the payment side of the scheme and another focusing on financing issues. Four existing mechanisms could possibly provide a legal basis for the payment side of a compromise EUBS: the multilateral surveillance procedure(Article 121(6) TFEU); fiscal as sistance in crisis situations(Article 122(2) TFEU); funds devoted to social cohesion(Article 175(3) TFEU); and the“flexibility clause”(Article 352(1) TFEU). Multilateral surveillance is a macroeconomic stabilization instrument for policy coordina tion. member states to follow Commission guidelines(Article 121(2)) in national economic policymaking. The same Article also impedes the Union legislator from introducing, by means of secondary legislation, sanctions other than those contemplated in Article 121(4). The fact that the multilateral surveillance procedure has been formulated on the basis of non-binding rules out its use as a basis for a future EUBS. This conclusion remains unaltered by the fact that Article 136(1), which makes reference to procedures covered by Articles 121 and 126, allows for the adoption of“measures specific to those member states whose currency is the euro”, given that the scope of Article 136 is intimately linked to that of the multilateral surveillance procedure(Repasi, 2017). The second is Union financial assistance to member states, which allows the Union to pro vide loans to member states“in difficulties or seriously threatened with severe difficulties caused by[…] exceptional occurrences beyond its control”(Article 122(2) TFEU). The clear connection to economic and financial problems inherent to this provision would seem to make it an adequate legal basis for a EUBS, particularly if the system were to be structured as an equivalent scheme. A closer look, however, reveals two important shortcomings. The first is that it only covers actions addressing crisis situations(Repasi, R. 2017). 10 The second is that the implementation of automatic trigger mechanisms would not be compatible with the fact that financial assistance must be granted on a case-by-case basis(Beblavý and Lenaerts, 2017). 11 10 This paper sets this legal base would be adequate for an equivalent EUBS with a trigger of> 2 of unemploy ment rate, which would only permit to activate the stormy day fund in our proposal 11 The Council nevertheless has broad discretion to define what constitute economic difficulties. It is also possible to think about ex ante conditions and not only ex post, which would facilitate the application of this Article. 14 www.fes-madrid.org

Sebastian Dullien and Daniel Pérez del Prado Some studies have concluded that the legal framework of funds devoted to social co hesion and, particularly, Article 175(3) could provide a legal basis for establishing an equivalent EUBS(Ferrante, 2016; Ministero dell’Economia e delle Finanze, 2016; Ministero dell’Economia e delle Finanze, September 2016a). The first paragraph of this Article states that the Union shall support the achievement of social cohesion and other objectives“by the action it takes through Structural Funds”. Of greater interest is an affirmation in a sub sequent paragraph that“specific actions necessary outside the Funds[…] may be adopted by the Council acting in accordance with the ordinary legislative procedure”. The authors of studies such those cited above hold that“specific actions” would cover everything needed to implement a EUBS. However, others who rule out the possibility of pursuing this option point out that given their focus on macroeconomic stabilization in times of crisis, equiva lent unemployment insurance schemes cannot be considered appropriate mechanisms for reducing social and economic disparities related to social cohesion(Repasi, 2017). The final possibility, upon which there has been greater consensus, is the“flexibility clause”. Article 352(1) spells out a number of limitations regarding its application. The first is that actions undertaken by the Union must respond to“objectives set out in the Treaties”. Of particular interest in the context of the creation of an EUBS are the objectives outlined in Article 3(3) of the Lisbon Treaty, which calls for establishing“a highly competitive social market”,“aiming at full employment and social progress” and promoting“social justice and protection”,“economic, social and territorial cohesion” and“solidarity among member states”. 12 The second is that actions be necessary“within the framework of the policies de fined by the Treaties” – in other words, any actions undertaken on the basis of this article must involve shared competences. In this context, it can be easily argued that the Euro pean level is the most appropriate level on which to compensate asymmetric shocks. 13 The third refers to situations in which the Treaties have not provided the necessary powers. The reference to“policies defined by the Treaties” infers that Article 352(1) may be invoked in instances that the Treaties do not provide a legal basis for actions needed to be taken. As none of the other three mechanisms analysed here would provide a clear legal basis for a EUBS, the flexibility clause appears to offer the only viable solution. Another factor that must be taken into account is an opinion issued by the Court of Justice of the European Union to the effect that Article 352 cannot be used as a basis for any new provision that would constitute an implicit amendment to the Treaty. 14 This effectively means that the 12 These objectives have to be read, among others, in conjunction with Article 9 TFEU, according to which in “defining and implementing its policies and activities, the Union shall take into account requirements linked to the promotion of a high level of employment, the guarantee of adequate social protection, the fight against social exclusion, and a high level of education, training and protection of human health”. 13 This is clearly linked to the principle of subsidiarity(Article 5(3) TEU). 14 CJEU, Opinion 2/94[1996], ECR I-1759, para. 30. www.fes-madrid.org 15

How to stabilize the euro area economy without creating political discord... distribution of competences cannot be altered and constitutional saving clauses must be respected. Germane to this issue is Article 153(4) of the TFEU, which states that provisions adopted pursuant to that particular Article“shall not affect the right of member states to define the fundamental principles of their social security systems and must not signifi cantly affect the financial equilibrium thereof”. 15 Furthermore, as a regulation establishing a EUBS on the basis of Article 352(1) TFEU would require the unanimous approval of the Council to be adopted, some member states could decide to veto the measure out of a desire to safeguard their constitutional clauses. Nevertheless, it must be remembered that any new initiative of this kind pursued on the basis of Article 352 would be subject to the“no bailout clause” embedded in Article 125(1) TFEU . Under this clause and according to the judgment handed down by the CJEU in the Pringle case, 16 any transfers of funding from the EU to member states not explicitly foreseen in the Treaties can only be justified if they foster some type of needed structural reform. In the case of social protection, the EU may grant financial assistance provided that it stimulates member states to implement reforms or new mechanisms related to their labour markets or unemployment protection systems. Three components of a number of EUBS models proposed to date could possibly provide the type of justification referred to above: experience rating, clawback provisions and minimum requirements for activation policies(Beblavý and Lenaerts, 2017; Repasi, 2017). Although the compromise scheme we propose here does not feature all of these elements, we are confident that given its national compartment component and dynamic claw-back provision, it would reduce moral hazard issues to a minimum and improve stabilization. Moreover, it would not offer national governments incentives to avoid implementing need ed structural reforms and can be deemed as not violating Article 125(1). It is also structur ally flexible enough to accommodate the incorporation of mechanisms supporting other facets of active employment policy such as experience rating and minimum requirements, should such adjustments be called for. The financing side of the scheme could be approached in two different ways. The first would be to set up a specific line in the EU general budget for Member State contributions and the second would be to create an external fund for this purpose. In terms of using a specific line for a EUBS within the general EU budget, Article 311(2) TFEU contemplates two kinds of funding sources:“own resources” and“other revenues”. Whereas funds in the first category are primarily intended to finance items within the 15 This is linked to the principle of proportionality(Article 5(4) TFEU). 16 CJEU Case C-370/12, Pringle, ECLI:EU:C:2012:756, paras. 130-136. 16 www.fes-madrid.org

Sebastian Dullien and Daniel Pérez del Prado general budget, those in the second may be used to finance specific purposes. Member contributions to the EUBS would fall into the category of other revenues, which offers two clear advantages: the regulation of other revenue is more flexible and such funds may be devoted to specific purposes. New contributions should be created and defined by the same legal instrument estab lishing the payouts and other aspects of the scheme they are intended to fully finance. Financial contributions to the scheme by EU member states, which are additional to the contributions they make under the Own Resources Decision 17 must, in principle, conform to the rules set out in Article 311(3) of the TFEU, which states that the Council shall, acting unanimously under a special legislative procedure and after consulting the European Par liament, the Economic and Social Committee and the Committee of the Regions, establish new categories of own resources or abolish existing categories that shall not enter into force until approved by the member states and in accordance with their respective consti tutional requirements. We have identified instances that could serve as precedents in which certain agencies seeking to establish new additional contributions(one example being the European Bank ing Authority) have avoided this procedure by using another Article as a legal basis. How ever, given that unanimous approval would likely be required for the establishment of additional contributions related to a new EU function, the previously mentioned Article 352 of the TFEU would provide the best legal basis for the additional contributions component of the EUBS(Repasi, 2017). On the other hand, contributions from member states may also be used to finance an ex ternal fund. The European Social Fund, the mission of which according to article 162 is to “improve employment opportunities for workers in the internal market and to contribute thereby to raising the standard of living”, is an example of a pre-existing EU structure that could play an important role. The main problem is that to date the ESF has been used ex clusively as a conduit for active employment policy initiatives. Giving it a passive employ ment policy function would require a major reform, the incorporation of substantially new components and probably a Treaty amendment as well. It would nevertheless be interest ing to explore the potential synergies between a future EUBS and this fund, which could possibly develop active employment policy initiatives that complemented the functions of the unemployment benefits scheme. 17 2014/335/EU, Euratom: Council Decision of 26 May 2014 on the system of own resources of the European Union. OJ L 168, 7.6.2014, p. 105-111. www.fes-madrid.org 17

How to stabilize the euro area economy without creating political discord... In any case, there are other – and perhaps easier to implement – possibilities that could be contemplated. Member states could sign an international treaty to create an intergovern mental organization, which, operating under public international law, provided access to financial assistance within the EUBS framework. The European Stability Mechanism(ESM) is an example of this kind of arrangement. Using Article 352(2) TFEU as a legal basis, it would also be possible to create a Union agency with a distinct legal personality, the bud get of which could function as a fund. Last, but not least, a subset of member states acting on their own could implement the scheme. This option would not require the unanimous approval of the Council and could be accomplished either by means of an international agreement accorded by participating member states(a solution rather similar to one mentioned above, but broader in scope) 18 or enhanced cooperation as contemplated in Article 20 of the TFEU. The first of these options is supported by the CJEU’s ruling in the Pringle case that mem ber states may conclude international agreements in areas in which the Union does not have exclusive competence. Certain conditions, however, would need to be satisfied. Such agreements must not modify Primary Law, be in compliance with Primary law and Second ary law, not encroach upon exclusive Union competences or shared competences, only be concluded if Union enhanced cooperation has failed or is likely to fail and must not circumvent Union legislative procedure or the Treaties. The greatest theoretical obstacle to pursuing the second option is the stipulation in Article 326(2) TFEU that enhanced cooperation must not undermine either the internal market or economic, social and territorial cohesion. However, a number of authors have concluded that this injunction would not constitute a barrier to the creation of a EUBS, which would not impede the social cohesion of the Union but only strengthen cohesion between participating member states(Repasi, 2017). Conclusions The compromise EUBS presented in this paper would be an equivalent scheme financed by contributions from member states that makes payments to national governments rather than individuals. As envisioned, each participating country would pay 0.1 per cent of its GDP a year into the scheme. Eighty per cent of its pay-ins would be deposited in a desig nated national compartment and the remaining twenty per cent would be pooled into a common stormy day fund reserved for dealing with severe shocks. 18 As in the case of the ESM, such an arrangement could be open to, and include, all member states. 18 www.fes-madrid.org

Sebastian Dullien and Daniel Pérez del Prado Pay-outs would be made whenever a country’s unemployment rates rose more than 0.2 percentage points above its average rate for the previous five years. The trigger for this process has been set intentionally low to avoid problems associated with higher thresh olds. Money for pay-outs would be drawn from national compartments. If a given country were to be hit by a very severe shock(defined as a rise in its unemployment rate of 2 or more percentage points) it would be eligible to receive additional payments from the scheme’s stormy day fund. Each participating country would be allowed to run a cumulative deficit in its national compartment of up to 5 per cent of its GDP. In the first instance, this deficit would be financed by loans from other national compartments. In the event that all national com partments should be depleted, the scheme would replenish funds as needed by borrowing in financial markets. To cover this contingency, the scheme would have the mandate to issue bonds backed by future contributions as collateral. Dynamic, risk-based adjustment of contributions would be built into the system to dispel concerns regarding the possibility of permanent transfers. This scheme demonstrated significant stabilization potential with minimal net costs to the net contributors under simulation conditions. In terms of the legal basis upon which this scheme should be created, a thorough analysis of the possibilities indicates that from a practical perspective the“flexibility clause”(Ar ticle 352 TFUE) would be the best choice. Nevertheless, it must be remembered that any new initiative of this kind would be subject to the“no bailout clause” embedded in Article 125(1) TFEU. Under this clause, any granting of financial assistance on the part of the EU to member states not explicitly foreseen in the Treaties can only be justified if it implies some type of conditionality. On the other hand, two approaches could be taken to the financing side of the scheme, which could be funded by means of either a specific line in the general EU budget or a dedicated external fund. Another workable alternative that would preclude the need for unanimous approval in the Council would be for a subset of member states acting on their own to implement the scheme. This could be done either by means of an international agreement accorded by participating member states or enhanced cooperation as contemplated in Article 20 of the TFEU. www.fes-madrid.org 19

How to stabilize the euro area economy without creating political discord... References Ÿ Ÿ  M. ; Gros, D.; Maselli, I.:“Reinsurance of national unemployment benefit schemes”, CEPS working document, 401, 2015. Available at: http://www.ceps.eu/node/9923. Ÿ Ÿ  M. ; Lenaerts, K.: Feasibility and added value of a european unemployment benefits scheme, Brussels, 2017. Ÿ Ÿ  S. : A European unemployment benefit scheme. How to provide for more stability in the euro zone, 1st edition, Gütersloh, Verlag Bertelsmann Stiftung, 2014. Ÿ Ÿ  S. :“Ten lessons from a decade of debating EUBS: robust findings, popular myths and remaining dilemmas”, Intereconomics: review of European economic policy(forth coming), 2017. Ÿ Ÿ  L. P.; Osterloh, S.:“Is a fiscal capacity really necessary to complete EMU?”, Paper presented at the workshop How to build a genuine economic and monetary union on 30 May 2013. Ÿ Ÿ  dell’Economia e delle Finanze: European Unemployment Benefit Scheme, Au gust 2016. Available at: http://www.mef.gov.it/inevidenza/documenti/Unemployment_ benefit_scheme_rev_2016.pdf Ÿ Ÿ  dell’Economia e delle Finanze: A European Unemployment Benefit Scheme: nine clarifications, September 2016a. Available at: http://www.mef.gov.it/inevidenza/ documenti/EUBS_chiarimenti_EN.pdf Ÿ Ÿ  dell’Economia e delle Finanze: Simulations on the European Unemployment Benefit Scheme for the 1999-2015 period, September 2016b. Available at: http://www.mef. gov.it/inevidenza/documenti/EUBS_simulazioni_EN.pdf Ÿ Ÿ  Beirat beim Bundesministerium der Finanzen: Zwischen Fiskalunion und fiskalpolitischer Eigenverantwortung: Zum Vorschlag einer europäischen Arbeit slosenversicherung, 2016. 20 www.fes-madrid.org

Friedrich-Ebert-Stiftung Oficina Madrid C/Covarrubias, 22 28010 Madrid Tel: (+34) 91 294 88 80 e-mail: info@fesmadrid.es www.fes-madrid.org Las opiniones expresadas en esta publicación no reflejan necesariamente los puntos de vista de la Friedrich-Ebert-Stiftung. No está permitido el uso comercial de los medios publicados por la Friedrich-Ebert-Stiftung(FES) sin el consentimiento por escrito de la misma.