Insurance-Type Cooperation Mechanisms under EU Law

In the economic policy domain, calls for insurance-type cooperation within the European Monetary Union (EMU) have frequently been made. Insurance-type cooperation relates to the debate about macroeconomic stabilization tools helping to absorb asymmetric shocks within EMU. Adopting a legal perspective, this article aims explores the scope offered under the current EU treaties in establishing such cooperation mechanisms. Among the broad variety of policy proposals, we focus particularly on both a permanent EU unemployment scheme and a shock-based insurance. We identify the potential legal basis for these insurance schemes highlighting differences in legal feasibility given their specific design. We also discuss the endowment of insurance funds setting out four different funding modes under the EU treaties or on intergovernmental basis. We show that legal scope for insurance schemes is limited. Fully-fledged unemployment insurance schemes are likely to overstretch the boundaries of the EU treaties. More narrowly designed, however, such schemes are rather likely to be feasible if set up as shock-based mechanisms, where the gravity of the economic shocks is significant.

component for a genuine EMU. However, given the existing reservations and obstacles to further deepening coordination mechanisms, ambitious treaty amendments appear (politically) unlikely and underscore the significance of effective application of the existing legal framework.
Against this background, the purpose of this article is to discuss the legal feasibility of insurance-type cooperation within the current framework of EU Treaties as one central element of the fiscal federalism model. More specifically, we shed light on insurancetype cooperation mechanisms that cushion large macroeconomic shocks and make the EMU more resilient overall. Section II briefly sets out the policy proposals surrounding insurance mechanisms and draws an analytical distinction between a permanent EU unemployment scheme and a shock-based insurance. Section III identifies the potential legal basis for these insurance schemes highlighting differences in legal feasibility given the specific design of the insurance. On that basis, Section IV discusses the endowment of insurance funds setting out four different funding modes under the EU treaties or on intergovernmental basis. Section V concludes.

II. Basic functioning of insurance models
Among the multiple potential designs of insurance-type cooperation, 12 we representatively explore the legal feasibility of two concrete modes of insurances. The main difference between them is that one is permanent in nature, that is, setting up a re-current flow of financial transfer based on a set of criteria even thoughin some modelsthe financial transfers are intended to even out over a certain period of time.
The alternative model is shock-based, that is, transfers occur only (and rarely) if certain thresholds reflecting a severe economic shock are met. insurance would replace the corresponding part of national schemes. The levels of the contribution and of the benefit should represent a relatively low common denominator between the rules of the various national schemes. The insurance would focus on short-term unemployment and would, for example, be paid only for the first six months of unemployment; and the amount would represent 40% of the previous reference wage. Each Member State would be free to pay out a higher or longer unemployment benefit on top of this European unemployment insurance. Crucially, this basic European unemployment insurance would help EMU Member States to share part of the financial risk associated with cyclical unemployment. Every month national authorities would send to the European fund the basic contribution from all their employed workers. Likewise, every month the European fund would pay to the national authorities an amount corresponding to the sum of all the basic European unemployment benefit payments to be made that month in the country. The overall volume of such a basic European unemployment insurance scheme would be around 1% of GDP, mainly depending on the exact parameters such as duration and level of the benefit or the eligibility conditions. This model stands for an unconditional, permanent insurancevariations can particularly modify the eligibility criteria. 14 An alternative key variable, even though maintaining the idea of creating permanent financial flows, would be the output gap, i.e. the deviation between actual and potential output. It considers developments in the whole economy and its evolution is not directly influenced by the Member State's labour market institutions. However, potential output cannot be measured directly and must be estimated through sophisticated methods.
As a result, output gap estimations undergo a continuous revision process and the revision bias is considerable. 15 The second insurance model is distinct from the above mainly in that payments from the insurance depend on the occurrence of an economic shock. These models vary in the underlying threshold criteria. Payments could be bound to deviation of actual from natural unemployment rate, deviations of short-term unemployment rate or the ten-<http://europa.eu/rapid/press-release_SPEECH-14-455_en.htm>; a more sophisticated economic modelling of EU unemployment schemes is explored by ÁBRAHÁM ET AL., supra n. 9.
year average and deviations from the average output gap. 16 If the threshold is met, national unemployment insurances receive payments from the EU fund. The distinct feature to the above insurance schemes is its exceptional rather than permanent character. Thus, this design of unemployment insurance seems suitable for large shocks rather than small national shocks. Deductibles are proposed as a balancing tool, that is, deductibles are being based on actual long-term average spending on benefits for the short-term unemployed. Finally, country premiums should be differentiated according to risk in order to ensure a rough long-term balance between contributions and benefits for each country.

III. Legal basis for macroeconomic insurance schemes
There is a discrepancy in analytical depth between the economic policy debate and the legal analysis of insurance models. While the discussion of macroeconomic desirability of an insurance scheme has been active for quite some time, producing a variety of different proposals, 17 no comprehensive legal analysis has been undertaken yet. 18 In the following, we will identify potential legal basis and constraints for the above models of insurance highlighting differences in legal feasibility of the various designs of insurance schemes.

A. Social policy issue
In principle, the applicable legal basis for EU measures depends on the subject area in which the EU intends to become active. In the case of an unemployment scheme aiming primarily at performing a macroeconomic function as discussed here, this does not seem to be a straightforward issue, as it relates to both governance of national EU unemployment schemes and a macroeconomic instrument to smooth business cycles.
Generally, in a case where the Union's legislative intention allows for more than one legal basis to be applicable, the choice of the appropriate legal basis depends on where 16 See, in particular, BEBLAVÝ, GROS & MASELLI, supra n. 9; for the various concepts, see EUROPEAN COMMISSION, supra n. 12.
17 See supra n. 9 above. See also the assessment provided by EUROPEAN COMMISSION, supra n. 12. 18 There are, however, some valuable analyses, e.g., MATTHIAS  In the absence of harmonization at EU level, it is in principle for each Member State to determine the conditions for insurances under a social security scheme and the entitlement to benefits under that scheme. As regards EU legislation on unemployment schemes, so far and apart from the issues of establishment, services and, in that context, supervision, the Union's two main approaches have been coordination (rather than harmonization) and negative integration (prohibition of various types of discrimination), based on Article 48 TFEU (which allows notably for coordination) and According to Article 153 (1) TFEU, the EU shall "support and complement the activities of the Member States". Further, Article 153 (2b) TFEU allows the EU to set minimum requirements by means of directives, among others in the area of social security and social protection of workers (Article 153 (1c)). While EU competences in social affairs have been gradually expanded over last Treaty changes, 21 a number of hurdles limit the EU's scope of manoeuvre in implementing an EU-wide unemployment scheme.
First, Article 153 (4) TFEU imposes a general restriction in the EU's exercise of its competences under this provision by requiring that any measure taken by the EU must not affect the Member States' freedom to determine the fundamental principles of their social security systems. 22 Such measures must not affect these systems' financial equilibrium either. Considering that unemployment rules are at the core of social security policy, any measure in this field would concern a fundamental pillar of national social policy. Further, the harmonization of the replacement, duration and conditions of eligibility possibly has significant financial implications for national security systems. 23 Thus, the design of the insurance scheme matters. In particular, the EU would need to introduce the scheme on the basis of the smallest common denominator of national insurance schemes and avoid harmonization and minimize financial impact rather than requiring substantial changes to national schemes. Third, Article 153 (2) b TFEU only allows the setting of "minimum requirements" and must be interpreted in conjunction with the supporting and complementing functions referred to in Article 153 (1) TFEU, which implies an overall restriction on EU harmonizing activity. 25 Any attempt to replace national unemployment schemes in their entirety must run counter to this norm. In turn, partial replacement of national systems may be compatible to the extent that the overall level of protection for the unemployed remains the same, while the EU would finance part of the replacement rate (which then still has to comply with the "financial equilibrium requirement" mentioned above).
In sum, a genuine unemployment insurance implemented on the EU level is not likely to be grounded on Article 153 TFEU given the high degree of the Member States' room for manoeuvre in the area of social security. After all, even if the ultimate and specific design and purpose of the insurance scheme matters, some general conflicts of such instruments with this provision can be identified. Given obvious incompatibilities with Article 153 TFEU minimizing the harmonizing elements may be an alternative with a view to complying with Article 153 TFEU. In turn, reducing the insurance scheme to the smallest common denominator of national insurance schemes would necessarily lower the desired macroeconomic effect. Finally, designing the insurance as shockbased rather than permanent might shift the legal basis and could trigger an application of Article 122 TFEU rather than Article 153 TFEU. This would then be in line with the second basic insurance model scheme described above. Designed as unemployment scheme, only fulfilment of significant employment threshold (e.g. large rise in shortterm unemployment) could then trigger financial transfers.

B. Emergency clause
In order for Article 122 (2) TFEU to serve as a legal basis for an insurance scheme, the entitlement to receive payments under the scheme would need to be tantamount to "severe difficulties caused by natural disasters or exceptional occurrences beyond its control". The relevance of this norm during the euro crisis gave rise to a controversial 25 BENECKE, supra n. 22, at para. 7. interpretation of its terms as well as its relationship to the no-bailout clause in Article 125 TFEU. 26 First, in relation to the requirement of "exceptional occurrences", there is considerable debate on whether this term not only covers obvious cases such as social unrest or foreign policy turbulences, but also extends to solvency issues due to public debt resulting from a financial crisis. 27 During the crisis, the controversy extended to whether the difficulties of Greece, Ireland, Portugal and Spain were caused by such occurrences. 28 According to a narrow interpretation of the norm, public debt does not qualify as "exceptional occurrences", even if it has also been caused by a global financial crisis. Also, given that "natural disasters" and "exceptional occurrences" are mentioned in the same breath, some conclude that Article 122 (2)  In the Council's view, this implied that the debt crises in Greece, Ireland and other eurozone states, which the EFSM was designed to alleviate, were entirely or largely caused by the 2008 recession. Thus, the deterioration that the regulation mentions must presumably have been unforeseen and sudden. The Council considers the states' debt problems to be a direct byproduct of the 2008 downturn. Critics counter-argued that there were no unforeseen circumstances. Fiscal management and chronically high deficits were a prominent feature of the Greek economy for several decades. 28  compromising view, however, distinguishes between the debt being either the result of an unsustainable budgetary policy or the result of the impact of a general financial crisis. 31 Considering the degree of flexibility granted by Article 122 (2) TFEU, and considering that every situation requires a case-by-case examination, a country's own responsibility in critical circumstances can be taken into account when determining the conditionality necessarily tied to financial assistance. 32 From these legal standards, one can infer guidance on how the insurance schemes considered above should be considered in light of Article 122 (2) TFEU. The exceptional nature of the situation in which financial assistance may be granted disqualifies any scheme that establishes a permanent transfer system, irrespective of the exceptionality of the economic situation. Unemployment schemes falling within the first model described above do not meet the standards of Article 122 (2), for they simply replace recurring elements of an insurance scheme that is responsive to usual business cycle fluctuations. Variations in business cycles cannot be considered exceptional within the meaning of Article 122(2) TFEU. By contrast, shock-based adjustment mechanisms may be designed in a way that would reflect the requirements of the norm. Most importantly, both the characteristics of the criteria and the applicable threshold would need to be chosen with a view to reducing potential payments under the insurance to a level that can arguably be considered as exceptional. The frequency by which support is triggered must plausibly be due to significant crisis effects. Unlike under the ESM, the exceptional impact triggering payments under the insurance can be limited to the country concerned and must not necessarily be of a euro-wide dimension. By way of economic assessment, the representative character of the respective criteria to measure the severity of economic shocks (e.g. short-term unemployment, output gaps) must be determined. For example, one study calibrates the trigger of the insurance to a "tornado scenario" (alluding to the "natural disaster" occurred in the period 2000-2012. 33 Without prejudice to a case-by-case analysis, three to four annual incidents leading to insurance payments hardly appear to be exceptional. However, the study highlights that ultimately it is a question of designing the insurance scheme and calibrating the thresholds in a way that is compatible with the standards of Article 122 (2) TFEU.
Further, it must be assured that the payments have not been caused in a foreseeable fashion by the recipient country in order to remain within the wide definition of exceptional occurrences referred to above. Reference can be made to criteria related to the foreseeability of the consequences. Generally, one may infer guidance from past applications of this norm related to fiscal conduct. While a permanently high deficit will almost automatically lead to increased market pressures, given the (more or less) causal relationship between deficits, debt and interest rates, the elements causing a short-term unemployment rate or output gaps are much less clear. This finding can also be brought into line with the Court's reasoning on the ESM in Pringle. The Court identified two obstacles to the use of that provision as a legal basis for an EU-based ESM. A first problem was that the ESM establishes a permanent mechanism, whereas the Court found the requirement of "exceptional occurrences" to 33 BEBLAVÝ, GROS & MASELLI, supra n. 9, at 27; they use as criteria the deviation of short-term unemployment from its long-term average. imply assistance on an ad hoc basis only. 34 Under the ESM, the ad hoc nature of the support was ensured through a formal decision of the Council rather than automatically triggering financial support. Similarly, the trigger of the shock adjustment mechanism could be made dependent not only on the threshold criteria, but also on a subsequent affirmative decision by the Council, accounting for the explicit ad hoc character required by the Court in Pringle. That such assistance is organized via a permanent (but shock-dependent) mechanism is irrelevant so long as assistance is granted in specific (emergency) instances and only for the duration of the (threat of) serious difficulties. 35

C. Compatibility with the no-bailout principle
Any financial assistance must be assessed in respect of the no-bailout clause set out in Article 125 TFEU. There has been a long-standing debate about how this norm should be interpreted. 36 In Pringle, the ECJ substantiated the terms of this norm by finding that this clause was "not intended to prohibit either the Union or the Member States from granting any form of financial assistance whatever to another Member

IV. Fiscal capacity for insurance scheme
Endowing the insurance fund raises the question on EU fiscal capacity. 45

A. Revenues through EU tax
The current treaty may offer a legal basis for the introduction of certain taxes. 46 In specific fields, the treaty provides for tax measures, such as Article 194 (3)  difficulties. 63 However, this norm has been applied in a manner acknowledging that "it is for the Council to decide whether to grant a loan or appropriate financing facility, its average duration, its total amount and the amounts of the successive instalments". 64 Thus, there is ample scope for manoeuvre for the Council to determine the appropriate method of support. This flexibility led the Council under Articles 143 and 352 TFEU to establish a medium-term financial assistance facility, enabling loans to be granted to Member States. Article 143 TFEU as the norm allowing financial assistance. 66 This systematic connection between the two norms justifies applying similar standards.
Furthermore, in light of the conditionality requirement referred to in Article 143 (2) TFEU 67 and the significance of conditionality as a requirement for financial support under Article 125 TFEU, 68 one might impose conditionality to loans under the insurance scheme as well. In practice, this would imply that shock-induced recipient countries would be subject to conditionality, which would set out the policy measures for which payments should be used, with a view to maximizing the macroeconomic smoothing effect of the support. However, financial assistance granted under Article 122 TFEU is not necessarily limited to loans. Similar to the overall mechanism of EU funds, financial assistance can be dispersed as non-repayable grants.

D. Contributions on intergovernmental basis
Considering (political) obstacles in extending the EU's scope of fiscal activities, it seems possible that an enhanced fiscal capacity could be funded more easily through direct national transfers. An enaction based on Article 136 (3) TFEU does not appear feasible, as the insurance primarily aims at balancing asymmetric business cycle shocks that do not seem to be "indispensable to safeguard the stability of the euro area as a whole". However, the crisis context has given sufficient examples of how Member States might put in place a mechanism providing financial support. 69  Indeed, economic policy still remains within the competence of the Member States (Article 4(1), 5(2) TEU). 70 Member States can adopt measures in this field, as long as the competences of the Union are not infringed. 71 The conduct of economic policy inherently enshrines the right to identify and implement specific measures and priorities according to a Member State's preference, and given the country-specific state of the economy. If an insurance scheme were based on economic policy as genuine Member State competence, there would be no conflict with competences assigned to the EUa conflict that would prevent Member States from establishing a fund outside the EU legal framework. If Member States are allowed to act outside of the EU framework, then Member States can also spend outside of the EU budget. 72 This strategy would be consistent with an intergovernmental model for the management of the Euro crisis, which has stressed the centrality of national governments (in the European Council) and their freedom to act through agreements outside EU law, rather than the centrality of the EU institutional machinery and the potentials of EU law to address the crisis. 73 Thus, an insurance created as an economic policy instrument on intergovernmental basis could generally be funded by Member States that would also regulate and govern its operational setup. Inconsistencies of such a fund with Article 125 TFEU, as had been at stake both with the ESM and the EFSF, 74 can be avoided if the specific design of the scheme complies with the requirements under this rule discussed above.

V. Conclusion
The current "surveillance model" of European economic policy coordination is likely to persist over the next years. Given the lack of political impetus in pursuing deeper integration towards the "fiscal federalism model", Member States will continue to have full fiscal competence and retain competence to conduct economic policy. In this scenario, the EU continues to be the "discipline enforcer" applying a number of fiscal rules and the existing budgetary and economic surveillance system. While introduction of fiscal federalism generally requires treaty amendments overcoming the current surveillance model, certain types of fiscal federalism may be feasible under the current legal framework. This analysis demonstrated this for an instrument that has been debated widely in the economic policy arenainsurancetype cooperation that responds to the frequent calls to strengthen the shock absorption capacities within EMU. It aims at stronger resilience within the EMU through shock absorption mechanisms requiring (at least some) fiscal capacities on the EU level and thus alluding to the logic of fiscal federalism.
Adopting a legal perspective, this analysis sought to highlight the space of manoeuvre offered by the current legal framework. While it is acknowledged that the absence of a clearly defined and institutionalized insurance scheme makes a (hypothetical) legal analysis more difficult, we identified some key legal requirements that any potential future insurance cooperation would need to take into account. In fact, the legal scope for insurance schemes is narrow and establishes a number of conditions on their design. Fully-fledged unemployment insurance schemes (even if limited to a kind of basic insurance) are likely to overstretch the boundaries of the EU treaties. In particular, the limited scope for EU competence in the field of employment policy would prevent any harmonization going beyond the smallest common denominator of national unemployment systems, which in turn would reduce the desired macroeconomic effect of such insurance scheme. More narrowly designed, however, such a scheme is rather likely to be feasible if designed as a shock-based mechanism where the gravity of the economic shock must be significant to bring financial transfers under the ambit of Article 122 TFEU. If properly designed as shock-dependent insurance, the requirements of Article 153 TFEU (for an unemployment insurance) would still have to be observed, while scope for an insurance tied to other key variables (e.g. output gap) would be wider albeit subject to the exceptionality requirement of Article 122 TFEU (let alone the economic disadvantages of output gaps as criterion 75 ).
The ban on bailouts imposed by Article 125 TFEU does not restrict the kind of transfers at stake, as they are not dispersed in times of acute financial market pressure, nor do they aim at ensuring financial stability in the euro area. Rather, such transfers would be another element in the array of historically grown financial assistance tools fostering social and economic cohesion in the EU. Regarding the establishment of the fund necessarily sidelining the insurance scheme, a practical first step would be the use of existing EU funds which could then be used as a nucleus for an extended funding 75 EUROPEAN COMMISSION, supra n. 12, at 7-8. Strategic Investmentsthe Commission seems to view this instrument as a first step towards an overall stabilization mechanism. Alternatively, intergovernmental initiatives for funding scheme may be a lawful alternative albeit fuelling the general tendency of establishing economic cooperation outside the EU legal framework.
What are the implications for the further trajectory of economic policy cooperation efforts within the EU? The surveillance model is likely to be the short-term avenue pursued by the EU institutions. In this vein, the recent Five Presidents' Report has stressed the use of existing instruments in implementing structural reforms. 76 On a subsequent stage, the Five Presidents' Report proposes the introduction of a fiscal treasury. 77 While such a far-reaching institutional novelty would certainly require treaty amendments and is realistically confined to the eurozone, this analysis has sought to highlight more limited models of introducing fiscal transfers aiming at macroeconomic stabilization, which may be more likely to be implemented within the existing rules. 76 Five Presidents' Report, supra n. 7, at 8.