VoxEU Column EU institutions EU policies Global crisis

Learning from mistakes: A ‘what if’ approach to assessing proposals for euro area reform

The policy discussion on euro area reform has entered a critical phase. This column, part of the VoxEU debate on euro area reform, attempts a ‘what if’ experiment based on the proposals in the recent CEPR Policy Insight. Focusing on the Greek case, it looks at the counterfactual case of such proposals having already been implemented at the outset of the crisis and examines their potential role in preventing the outbreak of the crisis or mitigating it once it was underway.

This column is a lead commentary in the VoxEU Debate "Euro Area Reform"

The policy discussion on euro area reform has entered a critical phase, and analytical contributions informing that discussion have become particularly important. Among these, the recent CEPR Policy Insight (Bénassy-Quére et al. 2018) stands out. It is an important contribution to the discussion which the European Commission kicked off institutionally with its own ‘roadmap’ proposals. Reacting to the Policy Insight, together with a number of other economists with diverse academic and policy backgrounds, I co-signed an opinion piece on a “Blueprint for a democratic renewal of the Eurozone” (Andor et al. 2018).

These two contributions share a number of common views on the state of affairs and the way forward. The starting point to both is the built-in institutional weaknesses of the euro area and the recognition that these have been only partly addressed to date. And both believe not only that the status quo is not tenable but also that it would be a mistake to settle for marginal changes. Where they differ is in their scope and intent. Bénassy-Quére et al. explicitly attempt to influence the ongoing Franco-German debate on euro area reform by presenting what could constitute an acceptable compromise. Andor et al. (2018) instead sketch out a bolder and broader direction that in the current political climate would seem to be beyond reach.

As a contribution to the debate engendered by Bénassy-Quére et al., I would like to offer some thoughts in the context of a ‘what if’ counterfactual. This consists of exploring ex post the impact of having Bénassy-Quére et al.’s reform proposals already implemented before the euro area crisis on preventing the crisis outbreak or mitigating it once it was underway. I will take the Greek case as my starting point and as main focus, in full knowledge that the characteristics of the Greek crisis are not representative or fully replicated in the other euro area countries which had to rely on emergency loans.1 Nevertheless, it is by examining the outlier Greek case that one can best draw conclusions on the adequacy of preventive and stabilisation mechanisms in place.

Counterfactuals are by nature difficult to sketch out and especially difficult to interpret in order to draw conclusions. In addition to the influence of the inevitable hindsight, one difficulty is in separating the counterfactual set up from actual historical events which would arguably not have happened in a policy environment vastly different as a result of having implemented certain policy proposals. Nevertheless, I will offer some reflections on the ‘what of’ question by looking at the potential impact of the different proposed reforms in Bénassy-Quére et al. (2018) on the factors which (i) triggered the crisis; (ii) made it worse than it could have been; and (iii) allowed contagion to other euro area countries. 

Starting with the initial trigger,at the root of the Greek economic and financial crisis was a triple deficit: a fiscal deficit, an external deficit, and one of confidence. In 2009, the country was running a public deficit in excess of 15% of GDP and a double-digit deficit in the external account. At the same time, the realisation that the deficit figures were severely underreported made markets wary of corrective policy pronouncements by the new Greek government. The situation was further exacerbated by the long delay of EU countries and institutions in the early months of 2010 in understanding that in the case of Greece, markets were effectively looking for a guarantee of no default. Market financing became gradually more expensive, leading to the in extremis creation of the bailout arrangements.

Would the crisis prevention proposals in Bénassy-Quére et al. (2018) have helped Greece avoid this outcome? A definitive answer to this question is obviously impossible. But following a simple expenditure rule guided by a long-term debt reduction target could have avoided the sharp increase in expenditures of the 2005-8 period, which the rules of the preventive arm of the Stability and Growth Pact (SGP) failed to properly monitor. This assumes effective monitoring, with the proposed independent national fiscal watchdog, supervised by an independent euro area-level institution, providing the early-warning signs which eluded Commission services in practice.

In such a situation, the huge discrepancy between reported and actual deficit data which contributed to the loss of market financing could have been avoided. Separating the role of an independent fiscal watchdog at EU level from that of the political decision maker would also have added to the credibility of EU institutions vis-à-vis the markets when assessing the Greek situation. 

Similarly, in a country which routinely used its domestic banking system in its public debt-financing exercise, the introduction of sovereign concentration charges for banks would have acted to effectively introduce a tighter market test when issuing debt and thereby reduced the large exposure of Greek banks to the sovereign, which undermined their position during the crisis. Simultaneously, a common deposit insurance would have at least tempered the continuous ‘bank jog’ of early 2010 which precipitated the need for financing from Greece’s EU partners and the IMF. 

In an unpublished companion note to Bénassy-Quére et al. (2018), Jeromin Zettelmeyer attempts such a ‘counterfactual’ for Greece, assuming all recommendations had been in place by 2001: the expenditure rule, banking union, sovereign concentration charges, the regime for debt restructuring, easy access to ESM liquidity, the rainy day fund. He runs simulations and concludes that if the government had followed the expenditure rule, Greece’s fiscal balance in 2009 would have been positive, as opposed to a 15% deficit; hence the crisis would have been averted. 

If, however, the government had ignored the expenditure rule but had still issued junior bonds, he believes one would have observed an earlier and less severe debt crisis, triggering an ESM programme with higher chances of success. Finally, if the Greek government had disregarded both the expenditure rule and the junior bond issue, there would likely have been a bailout in 2010, but on condition of debt reprofiling, followed by debt restructuring in 2012, again with higher chances of success.

These conclusions are reasonable; they point however to the inherent difficulty of such counterfactuals, especially in terms of how much of the initial environment they assume to be altered by ‘full implementation’ of the proposals. And should then one not also compare such a ‘what if’ with the alternative of full implementation of the existing institutional framework at the time? This would be relevant in judging the effectiveness of the expenditure rule versus simply following the SGP rules in the decade leading to the crisis. Had the government simply followed the SGP rules, with the EU institutions using the monitoring tools in place at the time more effectively, it is not unreasonable to suggest that the same crisis-aversion result could also have been the outcome in Greece.

Passing from prevention to mitigation, it is clear that once in the assistance programme, Greece suffered an enormous economic and social cost. The drastic fiscal consolidation, together with the internal devaluation to balance the external account, would under any circumstances cause a steep recession; however, the almost 30% cumulative drop in real GDP during this period is far in excess of what should have been necessary, even taking into account the adverse initial conditions. 

At the same time, the two additional programmes after the initial 2010 bailout are testimony to a series of failures, both external and internal. The former stretch from an initial programme design which was more a child of political realities amongst euro area countries than the result of robust economic design, to the delay in taking action on debt restructuring.2 The latter includes an important implementation deficit, the result of the inability or unwillingness of successive Greek governments and of the entire Greek political system to take ownership of the necessary reforms.

The proposals in Bénassy-Quére et al. (2018) do not address issues of programme design, or equally importantly the required political economy discussion which should have accompanied the bailout arrangements to ensure that the economic downturn is not converted into an economic and social collapse, with the associated political fallout and rise in populism. The second of the two contributions referred to in the beginning of this piece (Andor et al. 2018) comes closer to recognising the importance of these aspects. Proposals for a stronger macro stabilisation in the event of extreme shocks, such as a euro area-level unemployment insurance scheme, are important in this respect. Equally important are proposals for a new form of cohesion and convergence policy for countries with competitiveness and institutional challenges. Greece is a perfect example where investments in education, legal systems and infrastructure could have made a difference.

Of the proposals in Bénassy-Quére et al. (2018), however, one has a high relevance to mitigating some of the adjustment costs. It is the framework involvingsovereign-debt restructuring when solvency cannot be restored through conditional crisis lending. Reducing banks’ exposure to an individual sovereign (not only Greek banks in the case of government-guaranteed bonds, but also German and French ones), together with better stabilisation tools and a euro area safe asset, had they been in place at the time, would have helped euro area countries arrive faster at decisions on the inevitable debt restructuring of Greek official debt. The two-year delay in executing the private sector involvement – what Bénassy-Quére et al. call a “gamble of redemption”, but one driven by creditors rather than Greece – weighed on the adjustment effort by making market re-entry impossible and rendering the 2012 assistance programme unavoidable. 

Turning to the final issue of contagion, many of the proposals in Bénassy-Quére et al. (2018) would have helped avoid the Greek crisis spreading and becoming systemic in nature. A euro area fund assisting countries to absorb large economic disruptions, or a euro area safe asset offering investors an alternative to national sovereign bonds, had they been in place during the crisis, would have contributed to financial stability. Conversely, it is harder to state definitively the impact of the sovereign-debt restructuring proposals. ‘Fully implemented’, they would have indeed in principle rendered unnecessary the decisions taken at the October 2010 Deauville ‘walk on the beach’, which effectively precipitated Ireland and Portugal into assistance programmes.3

Full implementation, however, is not necessarily the right yardstick; we are always in a partial implementation environment, vulnerable to the political and market pressures of the moment. Even without Deauville, debt sustainability would in such crisis situations very much be ‘in the eye of the beholder’. The mere existence of the mechanism could instead well have led markets to assume that Greece was simply the first of many to fall. It is difficult to gauge whether rather than acting as a stabilisation mechanism, the existence of the mechanism would in fact force its broader than intended use.

In conclusion, the proposals in Bénassy-Quére et al. (2018) are a welcome contribution to that delicate dance between politics and economics which has always characterised attempts to reform the euro area. Had they been in place when the Greek woes triggered the broader euro area crisis, they would undoubtedly have made for a more robust system. It is unclear, however, whether by themselves they could have avoided the outbreak of the crisis or seriously mitigated its impact. The policy tools in place at any given moment are obviously critical; but for the eventual outcome, the defining difference may lie in the political economy of the situation. 


Andor, L et al.(2018), “Blueprint for a democratic renewal of the Eurozone”, Politico, 2 February.

Bénassy-Quéré, A, M Brunnermeier, H Enderlein, E Farhi, M Fratzscher, C Fuest, P-O Gourinchas, P Martin, J Pisani-Ferry, H Rey, I Schnabel, N Véron, B Weder di Mauro, and J Zettelmeyer (2018), “Reconciling risk sharing with market discipline: A constructive approach to euro area reform”, CEPR Policy Insight No. 91. 

Independent Evaluation Office (2016), The IMF and the Crises in Greece, Ireland, and Portugal: An Evaluation by the Independent Evaluation Office, Washington DC.

Mody, A (2014), "The ghost of Deauville",, 7 January. 

Papaconstantinou, G. (2016), Game Over – The Inside Story of the Greek Crisis, Papadopoulos Publishing.


[1] For an insider’s account of the Greek crisis, see Papaconstantinou (2016).

[2] While this is broadly recognised by academics and policy-makers alike, the IMF is one of the rare institutions that has attempted to provide an ex post analysis; see Independent Evaluation Office (2016). 

[3] The “Deauville decision” preannounced that in future, sovereign bailouts would require that losses be imposed on private creditors, and as a result drove spreads higher, arguably forcing the hand of Portugal and Ireland in requesting assistance programmes. For a more skeptical view on this, see Mody (2014).

1,470 Reads