Introduction
What is the impact of Greece's fiscal meltdown on the effectiveness of Europe's response? Using Perrow's (Reference Perrow1999) normal accidents theory (NAT), I argue that efforts to reduce the likelihood of a Greek default activated conflicting centripetal and centrifugal modes of governance. Greater centralisation in decision-making at the European Union (EU) level improves policy effectiveness because it addresses problems of contagion but it simultaneously raises the risk of overall failure by increasing diagnosis, coordination and compliance costs. Effectiveness refers to reduction in the likelihood of catastrophic failure, such as the Greek government's bankruptcy, by helping Greece to meet its financial obligations.
The crisis has exposed fundamental weaknesses in EU economic governance, threatening the stability of eurozone economies. Greece is one of the triggers to the broader EU financial crisis and, although it is not the only country being bailed out, it is an important barometer about the dos and don'ts of bailouts, especially as it affects “voluntary haircuts” to private sector investments (De Santis, Reference De Santis2012). In this sense, it serves as the catalyst to expose the institutional weaknesses of European policy-making. Lessons from the way the Greek crisis is handled are also used to address broader concerns with the future of the single currency.
I first discuss Perrow's NAT argument. Although it is framed in terms of accidents in large technical systems, Perrow's argument can be adapted to financial crises as well, linking the literature on crisis management to Economic and Monetary Union (EMU) (Lo, Reference Lo2009). In the next section, I explore interactions between Greek and EU authorities to highlight the degree of effectiveness in the way that the Greek crisis has been managed. Three episodes are explored: the first bailout in May 2010, the mid-term fiscal strategy (MTFS) in June–July 2011 and the second bailout in February 2012. Implications are drawn for theories of delegation, intergovernmentalism and the future of EU crisis management.
Normal accidents and policy response
In his study of accidents in systems utilising high-risk technologies (nuclear plants, weapons systems, chemical plants), Perrow (Reference Perrow1999) explains why some systems are more prone to accidents than others. He concludes that accidents are sometimes unavoidable because of flaws with the system's design; he calls them normal accidents. Normal accidents are rare systemic failures with catastrophic potential, such as nuclear meltdown or the collapse of a country's finances or banking system. Normal accidents are most likely in systems characterised by complex interactions and tight coupling. Unexpected interactions between two or more parts of the system may bypass the system's safeguards; if parts are also tightly connected to one another, damage cascades potentially bringing down a part or all of the system.
Complexity refers to the nature of interaction between units in a system while coupling refers to the strength of their connectivity. More complex systems tend to have a greater number of interactions between them, greater differentiation between units, and a relatively unrestricted flow of information. Bank interdependence through shareholdings or loans, the use of sophisticated financial vehicles, and government budgets are indicators of complexity. There are more feedback loops that, when they occur in unexpected ways, will impede control of systemic performance. Because the causal chain of events is elongated as units interact in unexpected ways, the possibility of mistakes or accidents increases. The greater the differentiation, the less likely it will be that units will comprehend each other's functions. The less units know about the whole production sequence, or conversely, the more they become preoccupied with their own tasks, the greater will be the likelihood of unexpected failures occurring. As the flow of information increases, the system becomes more chaotic and output becomes more unpredictable. Governments and banks are now bombarded with more information, but their ability to digest it might actually decrease as each unit focuses only on information it considers central to its task, neglecting warning signs of failure in other units.
Coupling refers to strength of connectivity between units or actors. The fixed exchange rates implicit in a single currency area, common interest rates and the 24/7 function of financial markets are indicators of coupling. Firstly, tightly coupled systems contain time-dependent processes. Problems cannot wait until they are attended to. Planned and unplanned interactions occur quickly while items must move quickly through the production process. For example, destabilising financial market turbulence requires immediate attention; it cannot be delayed or extended. Secondly, tightly coupled systems have little slack or excess idle resources. Quantities must be precise with little room for waste. Parts are not substitutable, so failure requires a shutdown of the process. The immediacy of feedback is an important element of coupling. Although all units of a system may receive feedback regarding their performance, the strength of their connectivity determines whether they are able to avoid damage.
Zahariadis (Reference Zahariadis2003) extends the argument to European integration. He applies it to different issues, while this article focuses on a single area, EMU. Greater complexity leads to decentralisation of decisions while tighter coupling leads to centralisation. The implication of this claim leads to the following hypothesis during a crisis:
Under conditions of greater complexity and tight coupling, the more policy is centralised to deal with coupling, the more ineffective it becomes in dealing with complexity.
Tight coupling raises the risk of contagion while higher complexity increases diagnosis, coordination and compliance costs. The forces behind the contradictory tendencies are laid out along with specification of costs that increase (or not) policy effectiveness. The argument's validity is greatly diminished if coordination or compliance costs are found to decrease rather than increase with more centralised decisions.
Higher levels of complexity make decentralisation more likely. As the number of interactions increases, the costs associated with diagnosing problems and coordinating a coherent response also increase. Data about fiscal outcomes, inflation and growth need to be collected and transmitted. Budgets, however, are not mere accounting devices but an evolving collection of complex deals between demands, rewards, practices, miscalculations and past behaviour by myriads of groups and policy-makers. Capital injections to save healthy banks may run foul of regulatory commitments, raising the need for coordination (Quaglia, Eastwood and Holmes, Reference Quaglia, Eastwood and Holmes2009). To prevent cross-border spillovers that adversely affect EMU, policies must be coordinated vertically, for example among functional partners such as monetary and fiscal authorities, and horizontally, for example across member states (European Commission, 2009: 59–60). Top leadership (the Commission) experiences information overload (budget figures have to be audited using common methods) because it now has to have more information to make timely recommendations to the Council of the European Union (“the Council”) when performance is deemed suboptimal. However, timetables differ across institutions and national governments, creating uneasy tension between short-term politics (electoral cycle) and long-term monetary stability (Dyson, Reference Dyson2009). Because standards of performance are not widely followed and may be contradictory (stability, sustainable growth and fairness), performance is likely to be considered perennially suboptimal and leaders will be constantly called upon to fix the system.
In such a task environment, adherence to strict system-wide targets raises significant compliance costs (Alves and Afonso, Reference Alves and Afonso2007). Objectives can be best accomplished when decision-making authority, especially when it comes to fiscal policy, is retained by national governments (Wyplosz, Reference Wyplosz2005). Each unit of the eurozone system (governments, banks, national agencies and so on) is given some autonomy to define and fix its own problems, leaving room for top leadership (the Council or European Central Bank, ECB) to intervene when failure or suboptimal performance reaches the threshold point of severe systemic damage. This is what happened during the banking crisis in 2007–08. Discretionary fiscal stimulus was released to shore up demand and ease economic hardship on a national scale while liquidity and guarantees were provided on a case-by-case basis to keep the financial system functioning (Quaglia, Eastwood and Holmes, Reference Quaglia, Eastwood and Holmes2009). The net effect is that centrally defined rule-based governance, such as the Stability and Growth Pact (SGP), has limited effectiveness under complexity because the problems afflicting each country and the solutions vary widely. The logic helps explain why SGP rules were reformed in 2005 to accommodate national political circumstances and needs (Heipertz and Verdun, Reference Heipertz and Verdun2010). As the Council readily admits, “a uniform medium term budgetary objective of [‘]close to balance or in surplus[’] imposed in some countries [that were] experiencing high nominal growth an undue policy stance […] rather, differentiated medium-term objectives are set for each Member State,” taking into account a different combination of deficit, debt and projected growth levels (European Commission, 2006).
When connections between units are strong, damage at one point may easily spread to others. There is little room for experimentation. Because the production process is invariant, contagion is the main problem; even minor damage has the potential in a short period of time to grind the whole system to a halt. For example, a possible sovereign default in a tightly coupled system, such as the eurozone, may lead to capital flight from several countries using the euro, not just the country experiencing bankruptcy because investors fear that other weak economies may also collapse, bringing down the value of the euro and with it the value of their investments (De Santis, Reference De Santis2012). Decisions by credit rating agencies to downgrade government bond ratings certainly caused a flight from some countries. Besides, “institutional investors, such as pension funds and insurers, are obliged by law or their own statutes to purchase and hold bonds with a certain minimum rating” (De Santis, Reference De Santis2012: 7). The 1997 Asian financial crisis demonstrated the tremendous speed of cascading damage across economies that were nowhere near as closely integrated as the eurozone (Haggard, Reference Haggard2000). Under these conditions, centralised decision-making and administration are the best option (Peters, Reference Peters2011). Rules are clearer and more precise and implementation is far speedier. There is greater coherence and political control in crafting an appropriate response to disturbances. Under conditions of tight coupling, centralisation makes it faster for systems to recover from errors (Perrow, Reference Perrow1999: 333). Centralisation means information is collected, analysed and implemented by a central unit, the Commission, Council or the ECB, rather than collected, interpreted and implemented separately by individual governments and national central banks.
Damage and policy ineffectiveness differ along the complexity/coupling dimensions, implying varying costs and benefits for different parts of the system. The real systemic threat comes from contagion; widespread bank failure in one country may bring down the entire financial system and its EU creditors. Nevertheless, the source of failure is likely to come from complexity as units may malfunction because of bad practices, corruption, or irresponsibility. The implication is that, under complexity, the damage may be small from the system's perspective but the cost of responding over time is high because of delays due to institutional friction. Institutional friction refers to eurozone costs associated with making a decision (coordination costs) and ensuring implementation (compliance costs) (Jones and Baumgartner, Reference Jones and Baumgartner2005). Ministers need to get together, agree there is a problem, define it and then decide how to act, by which time failure may metastasize. Because governments are accountable to different sets of national voters, they don't have incentives to speak with one voice. Creditors are more interested in addressing contagion by shielding their own financial institutions from Greek default, while debtors want to eliminate bad practices and corruption. Governments may agree on a plan that addresses both, but they have political incentives to implement different parts of it. The greater the degree of centralisation to limit contagion, the greater will be the extent of recipient resentment at having to “implement policies likely to be seen as dictates from Commission bureaucrats and self-interested foreign governments trying to protect their own banks, investors and export industries” (Scharpf Reference Scharpf2011: 31). Under these conditions, the political will to comply declines and so does policy effectiveness.
Tightly coupled, complex systems are torn between two opposing forces. They have to centralise and decentralise responses at the same time. Political conflict rises and so does the cost of response over time. Issues are framed and reframed in ways that address one set of problems or the other. In doing so, new actors with competing constituencies are mobilised to claim legitimacy and control (Daviter, Reference Daviter2011). A crisis increases urgency, frustrating the need for balance between effectively dealing with contagion and complexity. Governments may act quickly to save Greece from meltdown, for example, increasing centralisation and control, but doing so leads to higher compliance and coordination costs. Recipients do not have incentives to comply not only because they resent what they perceive to be external directives but also because austerity measures raise domestic political obstacles. Voters will resist implementation, fearing dramatic downgrades to their own welfare. The harsher the measures, the greater resistance will be. This is particularly true in tax collection administration because many taxpayers estimate the benefit (avoidance of bankruptcy) as outweighing the cost (paying a fine when caught) and are tempted to use non-compliance as an alternative source of financing operations (Bartolo, Reference Bartolo2009: 5–6).
Coordination also becomes a problem as frustrated creditors witness recipients trying to accomplish what is politically feasible instead of delicately sequencing measures to maximise effectiveness. For example, privatisation schemes may be delayed due to opposition by labour unions protesting against cuts to broader social safety nets and unemployment benefits (Kikeri, Reference Kikeri1997). Employee lay-offs may also lead to staffing shortages of agencies in charge of privatisation. Both elements give rise to incoherent responses and delays, ultimately limiting policy effectiveness.
The more systems tilt toward centralisation, the more they exacerbate problems with complexity. The more responses decentralise power, the greater will be the risk of ineffectiveness due to cascading damage. The temptation is to balance the two but these risks double the costs during crisis because of incoherence and delay. Indeed, the SGP has been characterised by both modes of governance: hierarchy when it comes to objectives (fiscal prudence and price stability) and negotiation (decentralisation) when it comes to actual implementation (Verdun, Reference Verdun2009; Heise, Reference Heise2008).
Coupling and complexity pose different kinds of problems. Tighter coupling limits time horizons and increases the risk of contagion while higher complexity increases diagnosis, coordination and compliance costs. Each problem pushes and pulls the system in different directions. In tight coupling, effective responses lower the risk of capital flight by centralising decision-making and decreasing bond spreads. Because the risk of contagion affects many units in the system, resources are pooled. As long as goals are widely shared, responses will be timely and uniform throughout the system and cost per unit will be relatively low. Complexity increases the need for intergovernmental negotiation and delay at the EU level because different units are allowed to articulate their concerns and implement responses according to their own pace and needs. The more decision-making is centralised at the EU level, the more effectively contagion is addressed. At the same time, diagnosis, coordination and compliance costs rise, increasing the likelihood of policy ineffectiveness.
Assessing the politics of response
How effective has the European response effort been? The tension between centralising and decentralising tendencies leads to a conflict between effectively addressing contagion and dealing with problems posed by complexity. The argument is probed in the context of three episodes of the Greek crisis: the first bailout in May 2010, the MTFS in June–July 2011 and the second bailout in February 2012.
Is it really a crisis? The first bailout
When the incoming socialist government of Pasok announced a revision to the 2009 budget deficit, it assumed it would be business as usual. George Papandreou, the Greek Prime Minister, engaged in the time-tested strategy of shifting blame to previous governments to minimise the political cost of taking unpopular decisions (Zahariadis, Reference Zahariadis2010).Footnote 1 He uncovered a budget shortfall of more than €16 billion and invited the European Commission to audit the books.Footnote 2 Uncertainty over Greek figures (expressed repeatedly by incoming governments and the Commission since 2005) resulted in a downgrade of Greece's sovereign debt by Fitch, Standard & Poor's, and Moody's driving up borrowing costs. On 12 January 2010, the European Commission formally rebuked Greece for falsifying public finance data, “systematic overspending, endemic tax evasion and persistently overoptimistic tax projections” (European Commission, 2010: 3). This caused further jitters in an already anxious market. Investors dumped Greek government bonds, further increasing the yield spread by close to 400 basis points above the German benchmark, leading to Greek cuts equalling €4.8 billion (Kontogiannis, Reference Kontogiannis2010).
Complexity made it difficult to define the crisis. Because EU members did not anticipate an attack on the euro, no mechanisms were put in place to diagnose such problems and guide the EU's response. Was it a national or an EU problem? The Greek prime minister and some EU officials insisted it was the former, leading to a decentralised response. In January 2010, Mr Papandreou emphatically declared: “The problem we have is home-made […] We need no bilateral loans, we have never asked for bilateral loans” (Elliott, Reference Elliott2010). At the same time, ECB president Jean-Claude Trichet echoed the sentiment: “Each country has its own problems. It [the Greek budgetary crisis] is a problem that has to be solved at home. It is your own responsibility.” There was political and economic gain in containing the crisis to Greece. It gave the impression of the government being in control over its finances, pacified angry taxpayers abroad and saved EU leaders from having to invent novel instruments of response. Besides, the ECB and several governments were still reeling from the fallout of having to inject billions of euros since 2008 to save the infected banks from toxic assets (Trichet, Reference Trichet2010).
Markets were unconvinced that Greece could do it all on its own largely because of the huge amounts involved and the government's poor finances. Papandreou then turned to his EU partners for support. As early as December 2009, EU Monetary Affairs Commissioner Joaquin Almunia said: “the problems in Greece are problems of the euro area” (Petrakis, Reference Petrakis2009). He saw the risk posed by the Greek problem as an issue of contagion, which should have prompted a centralised response. The Commission saw the danger but national leaders steadfastly refused. Ecofin's first response was that it was Greece's problem. As Featherstone, (Reference Featherstone2011: 202) observes, “it required Greece (under Article 126(9)) to cut its deficit and correct its divergences, thereby ‘removing the risk of jeopardising the proper functioning of EMU.’” While endorsing the Greek austerity programme, political divisions and the lack of details failed to pacify investors.
Some national leaders refused to openly admit that the problem was systemic. Instead, they were comfortable, as the hypothesis predicts, with the SGP's decentralising tendencies raised by complexity. Each country was responsible in fixing its own problems with public finances. The issue was particularly acute for German Chancellor Angela Merkel. Christian Democrats and her Free Democrat partners faced a tough election in the state of North Rhine-Westphalia on May 9, 2010. Merkel postponed making a decision on the rescue package for several weeks until panic in financial markets forced her hand. The drawn-out, derisive debate in the German press fuelled speculation that Germany wanted to avoid it all together. “Chancellor Merkel hesitated over the Greek bailout, and the whole issue stirred up public fears over the economy, unemployment and the stability of the euro,” said Manfred Güllner, head of the Forsa polling institute (Moore, Reference Moore2010).
Finally, EU leaders agreed, on 25 March, on a pledge to provide €25 billion (later raised to €45 billion) to Greece, if needed, via bilateral three-year loans from other eurozone member states and the International Monetary Fund (IMF). The idea was to prevent cascading damage by deterring speculation over impending default. Of course, the mere idea of a loan inspired more speculation because it pinpointed Greece's weakness and invited markets to call Europe's bluff. On a political level, Merkel initially insisted on market rates but on April 11 agreed to a maximum interest rate ceiling of about 5 per cent – below the 7.5 per cent the Greeks had been paying to that point but still hefty enough to pacify German voters. Papandreou and several EU leaders favoured a Europe-only solution, but Germany wanted IMF participation, fearing it would end up footing the bill. IMF involvement amounted to a perceived “historic” defeat for the euro (Manolopoulos, Reference Manolopoulos2011: 225), but the IMF's presence also served a useful political function, namely to diffuse responsibility, address a challenge in Germany's constitutional court, and deflect some blame from the impending pain away from Europe's “benevolent” governments to the Washington-based “villain.”
Several EU governments fiercely opposed an effective central (EU) response, fearing political backlash at home. Eurozone members would coordinate activities, but each government would retain veto over use of the facility. To provide loans, the creditors demanded additional cuts in Greek wages and pensions, affecting every aspect of the country's economic activity to an extent not seen since the civil war days of the 1940s. Confirming the hypothesis posed earlier, delay led to an uncomfortable balance between centralised decision-making and decentralised compliance.
On 23 April 2010, Eurostat increased its estimate of Greece's budget deficit to 13.6 per cent, triggering a rise of Greece's two-year bonds to 10 per cent. On the 27 April, Greece's debt was reduced to junk status. Unable to borrow on such terms, Greece informed its partners that it was activating the rescue package, invoking Article 122(2), which allows aid in case of emergency. The package (known as the Memorandum of Understanding, MoU) involved the gigantic amount of €110 billion over three years – €80 billion in bilateral loans and €30 billion by the IMF (2010) – monitored and disbursed quarterly on the basis of progress on specific indicators assessed by the IMF, the ECB and the European Commission (the so-called troika). The point was to temporarily help Greece generate primary surplus to meet obligations to external creditors at a reasonable interest rate; it is “the only possibility we have to ensure the stability of the euro” according to Chancellor Merkel (Hope et al., Reference Hope, Tait and Peel2010: 1). In return, Greece agreed to a series of painful measures designed to reduce its public sector, cut salaries and pensions, and increase tax revenues through VAT tax hikes on all goods and services in addition to more taxes on tobacco, gas and alcohol, and better collection processes. A member of the cabinet later admitted that he had “less than three hours” to read and assess the rescue package in its entirety before voting on it. Another, Michalis Chrisochoidis, later publicly admitted that he had not read it at all; yet he and others were asked to implement it for the next three years (Papadopoulos, Reference Papadopoulos2011)! Greek voters were stunned, confused and angry.
The Greek and EU responses were highly ineffective due to delays. Initially, EU leaders decentralised decision-making, letting the Greek government deal with its own problems. Article 125 explicitly prohibits national bailouts. Institutional friction delayed collective decisions and political cost amplified Greece's problem, fuelling the impression that the EU had to intervene. By the time it did, the threat of contagion prompted fears of a domino effect. A Greek default would also infect its creditors (largely French, German and British banks). The evidence supports the hypothesis’ prediction: the more the EU insisted on decentralisation, the higher the risk of contagion became, increasing the likelihood of a centralised response.
Failure, recrimination and weakness: the mid-term fiscal strategy
The haste with which the package was put together raised problems with coordination and compliance. On the one hand, Greek ministers (and MPs) were asked to support binding measures they had neither read nor understood. They couldn't possibly implement them effectively, let alone raise billions in new tax revenues with an unreformed tax collection system. On the other hand, there was political agreement (to an extent) on the need for austerity. The finance minister starkly stated: “In less than two weeks, a 9 billion-euro bond comes due and the state coffers don't have this money” (quoted in Featherstone Reference Featherstone2011: 203). Borrowing from foreign markets was prohibitively expensive so the only option was to accept the rescue plan.
The hypothesis states that increasing centralisation would make compliance more difficult. Evidence confirms that this was the case due to corruption and political cost. Despite the recognition that fundamental reforms were necessary, opposition parties dissociated themselves from the crisis, creating economic chaos and confusion. The communists and radical left (KKE and ∑ΥPIZA) called for general strikes, explaining the crisis as a problem for capitalists. Workers had nothing to do with it, and they should therefore not have to pay the price. How the country would pay its creditors or find money the day after bankruptcy remained a mystery for others to solve. The argument resonated well among the economically weak segments of the population who saw their incomes plummet by at least 15 per cent. The main opposition, New Democracy, sought to exploit popular discontent, figuring disaffected voters would eventually move to the right. The leader, Antonis Samaras, rationalised this contradiction: yes to the loan, no to austerity (calling instead for more pro-growth policies).
For its part, the government implemented many of the reforms but with an eye toward minimising political cost. Compliance costs rose dramatically and so did policy ineffectiveness. As a result, targets were systematically missed. For example, tax collectors went on strike, protesting wage and benefit cuts and reducing revenues by €4.5 billion below target (Hadjinikolaou, Reference Hadjinikolaou2011). Reforms of pension funds initially called for all 27 different funds to be merged into a single fund but socialist MPs insisted on exceptions, especially when it came to social security for those in parliament. Opening up closed professions led to sometimes violent opposition by affected groups so some liberalisation plans were postponed (for pharmacists and lawyers, for example) while others were implemented (truckers, for example). However, even those were crafted in a way that delayed resolution. For example, trucking would be liberalised but after 2013, implying that the measure could be quietly dropped when the government was no longer bound by the MoU terms! New taxi licences would be given – none had been since the early 1970s – but only based on strict population criteria. Because the number of taxis per person was already higher than elsewhere in Europe, the implication was that no new licences would be granted. It was a way to liberalise, and engage in, damage control.
Failing to form a coalition government with the conservatives, Papandreou reshuffled his cabinet in June 2011 to give new dynamism to his government, inviting the former competitor for the party's leadership, Evangelos Venizelos, to become finance minister. At the same time, the government submitted the MTFS plan to the IMF to amend the terms of the MoU. Because targets proved elusive, the new plan aimed to ease compliance by stretching fiscal consolidation to early 2015. It intended to save €28.3 billion (12 per cent of GDP) in order to reduce general government deficit from 7.5 per cent of GDP in 2011 (since revised to 9.6 per cent) to 2.6 per cent in 2014 (IMF, 2011). Despite Herculean efforts to streamline public finances – in 2010, Greece accomplished the highest reduction in general government deficit in the eurozone by 5 per cent of GDP – the government had simply come up short because of higher compliance and coordination costs.
EU leaders grudgingly accepted failure and agreed to the MTFS in an extraordinary European Council meeting on 21 July 2011. The package envisaged two additional items. First, the country would get €109 billion in new loans but under better terms (a low interest rate of 3.5 per cent with 15-year maturities). Second, private holders of Greek debt would take a voluntary “haircut” (or discount) on the nominal value of holdings by roughly 20 per cent through exchanging old for new debt or swapping Greek government bonds into debt with maturities up to 30 years (Baker and Toyer, Reference Baker and Toyer2011). More importantly, the ECB was instructed to continue providing credit guarantees to Greek banks and buying back government bonds if needed. As a result, ECB liquidity support to the Greek banking sector was projected to reach the gargantuan amount of €100.8 billion in 2011 (from €4.9 billion in 2006), amounting to 44.5 per cent of GDP (IMF, 2011, Table 3).
The deal addressed the problem of contagion by shoring up foreign and Greek creditors (government and ECB loans were not discounted). Some centralisation of decisions in the hands of the troika helped postpone default and thus avoided contagion. Experts were dispatched to Athens to form the Task Force, whose job was to diagnose and provide technical support with reforms. Progress would be monitored more closely, but the ability to deliver rested strictly with Greece. It was, after all, a democratic sovereign state although the terms of governance were increasingly written by its creditors. Decentralisation still made sense.
But evidence demonstrates that complexity still rendered policy ineffective because of a general lack of coordination of government activity and higher compliance costs. Strikes became more frequent and disagreement within the governing party and among political parties intensified as each blamed the other for failure. New taxes in Greece provoked more discontent, exacerbating compliance costs. The minister of finance demanded that reforms be quickly implemented, especially as they related to labour redundancies, privatisation and tax collection. All three entailed conflict with the unions. As part of the MTFS, the government had agreed to sell assets worth €5 billion in 2011, an amount it has since announced it will miss by €3.6 billion. Coupled with redundancies in general government (but not affecting employees in central government), it came face to face with powerful public sector unions, whose members were mainly Pasok voters. Taxes on property, social solidarity and others were imposed retroactively to the beginning of the year and given a new twist. Property taxes would be included in power bills, turning the power company into tax collector. The leader of the company's militant union, Nikos Fotopoulos (2011), declared with impunity: the power company “will not be used as a tax collection mechanism because it was built to serve not to blackmail the Greek people.”
Chaos: a Greek word to describe the second European rescue effort
Despite the centralising logic of new institutions and rules of economic governance at the EU level, a strong element of decentralisation, what Puetter (Reference Puetter2012) calls deliberative intergovernmentalism, prevailed. The shape, nature and pace of rescue remained firmly in the hands of national political leaders. While the Dutch and German governments insisted on haircuts to private sector holdings, a new parliamentary majority in Finland argued for more. The Greeks would have to put up assets as collateral to assure Finnish taxpayers that their money was safe. That was, of course, not included in the deal, complicating an already difficult situation. Similarly, the Slovak government resisted paying its share of the rescue loans; the matter was put to a vote of confidence and the government promptly collapsed. Amidst political conflict within and across national borders, power appears to have been redistributed among EU member states. The leadership role of the Franco-German partnership has been replaced by reluctant German paternalism. While the French President closely consults with Chancellor Merkel, Germany appears to often get more of what it wants.
The Greek government almost immediately sought to renegotiate the terms of the deal it had just made because it was clear that it would not meet its targets. Venizelos admitted that the problems of Greece were not just economic, they were mainly political, thus identifying the source of non-compliance. Papandreou argued that he could not take any new measures: “and this is something our creditors must understand” (Kroustalli, Reference Kroustalli2011)! The troika representatives were not impressed. Having assessed progress, they noted significant delays, unwillingness and indifference on the part of ministers to implement measures outlined in the MTFS. Although they agreed that the recession was complicating things, they insisted that it was a matter of coordination and political will. In one case, they encountered a minister who, responding to their complaints about delays, retorted: “the Memorandum is only indicative” (Kroustalli, Reference Kroustalli2011)!
Decentralisation was not working. Debtor and creditors were divided on what to do, exacerbating, as the hypothesis predicts, coordination problems. Some in the Greek governing party wanted to defuse public discontent by calling for new elections. Others feared that the results would not bring many of them back to parliament. In the meantime, political parties on the left openly called for insurrection to bring down the government while those on the right demanded new elections. Members of the eurogroup were equally divided on what to do. The heads of Germany and Holland pressed for more losses to be imposed on the private sector while France and the ECB feared that reopening the bond deal would spark renewed sales of shares in banks holding (not just Greek) bonds of peripheral countries (Spiegel and Peel, Reference Spiegel and Peel2011). Several days earlier, Germany's finance minister, Wolfgang Shäuble, alluded to bigger contributions by private bond holders amidst speculation by the Free Democrats, the government's junior partners, that Greece should be expelled from the eurozone. Banks were worried. Charles Dallara, head of Institute of International Finance, the global consortium of banks and investment houses, openly worried about the lack of EU leadership: the euro's successes “are being masked and undermined by parochialism and nationalism” (Beattie, Reference Beattie2011). President Barroso called for a new Eurobond, an option strongly opposed by Germany. The cacophony and hesitation among Europe's politicians made matters worse.
Rising spreads throughout the EU led to the realisation that damage reached the bonds of other governments. To put an effective end to contagion, a solution was found in the emergency summit on 26–27 October, which amended the 21 July package. Greece would get another €130 billion of fresh loans, €30 billion of which would go to prop up domestic Greek bondholders who would almost certainly become insolvent after the agreed 50 per cent haircut on private holders of Greek debt. While Greece would retain nominal implementation powers, EU advisors would be dispatched to assess compliance. In return, if all went according to the amended MTFS plan, the Greek government would bring its public debt ratio back to the 2009 levels (125 per cent of GDP) by 2021 while generating primary surpluses. Stressing centralised control, all eurozone governments agreed, among other things, to pass a balanced budget provision by the end of 2012 and subject their annual budgets to assessment by the Commission.
A more effective response to contagion exacerbated problems posed by complexity, leading to more political conflict. Greece was saddled with more debt and closer supervision in the name of effectively preventing its default and consequent collapse of Europe's banking system. Though nominally negotiating with its creditors, it is highly doubtful that the Greek government had any input into the bailout package given its dire economic condition. It had to accept imposed measures from external creditors. But the country had reached an impasse. Resistance to more austerity culminated in weekly general strikes and protests organised by two confederations of unions, ΓΣEE and ΑΔEΔΥ (private and public sector workers). More importantly, the government could no longer coordinate implementation of the programme it had just agreed. Several prominent MPs and former ministers declared they would not vote for impending labour reforms, charging that the party had strayed from its socialist roots (Kyriakidou and Neogy, Reference Kyriakidou and Neogy2011). This effectively deprived the government of a majority in parliament. At the same time, major government figures, including the Greek President, were heckled during the annual celebration commemorating the country's role in World War II. Greece was on the verge of a social revolution.
Fed up with policy ineffectiveness, disaffected by the prospect of a Greek referendum on the terms of the second bailout package, and humiliated by the Franco-German response, members of Pasok called for the Prime Minister's resignation. A compromise was crafted to vote on labour reforms, after which Papandreou agreed to resign. Initially, the opposition was reluctant to concede to a government of national unity, fearing the political cost. Eventually, Samaras relented, paving the way for a technocratic government headed by Loucas Papademos, the person intimately linked to Greece's entry to the eurozone while serving as governor of the Bank of Greece. While Greeks have generally welcomed the transition, political parties are positioning themselves for the day after general elections, which are scheduled to take place sometime in April.
Yet another rescue was agreed in February (and signed in March) 2012, which modified slightly the terms agreed in October 2011 (Castle, Reference Castle2012). Officially, it is the second package because the rescue agreed in July and October 2011 was never implemented. Besides agreement with private banks to accept a voluntary haircut of 53 per cent, the new bailout package includes much more centralised control of implementation. Specific measures are spelled out in greater detail, such as giving numerical targets for public employee reductions, prerequisites are given for laws in addition to specific timeframes, and agreements are signed with European (mostly German, Dutch and French) companies to take over implementation and compliance with the new terms in such areas as tax collection, health care restructuring, strategic planning, organisational reform and others. It is too early to assess effectiveness, but the argument made here implies even higher compliance and possibly coordination costs. Compliance costs may increase as resentment rises against even more visible outside control. Coordination might improve with the involvement of more skilled aid from abroad, but political conflict will rise as Greek politicians will surely attempt to shirk responsibility for any failures. Centralised control gives greater coherence to policies and their precise sequencing in order to maximise the chances of success. However, problems with the complexity of the task raise the strong possibility of failure. Although the package was adopted by the Greek parliament, political instability spells growing compliance and coordination troubles. Political parties are divided as to the desirability of the rescue and European and Greek officials privately express doubts over the ability of Greek administration to deliver on such a complex task (Donadio, Reference Donadio2012).
Centralisation, complexity and policy effectiveness
The Greek crisis has exposed structural weaknesses in EMU crisis management. When governments face a crisis, they tend to centralise decision-making to prioritise and maintain control of their response (Peters, Reference Peters2011: 77). But that is problematic because of contradictory forces. The tight coupling of the monetary system creates problems of contagion and necessitates centralisation of decisions and immediate responses. The complexity of interactions and possible spillovers to and from other areas of the economy promote decentralisation and experimentation with implementation (Boin et al., Reference Boin, ‘t Hart, Stern and Sundelius2005). Confirming the NAT hypothesis, the Greek case demonstrates that, under crisis conditions, greater centralisation in decision-making improves policy effectiveness in dealing with the risk of contagion, but it poses high diagnosis, compliance and coordination costs raised by complexity. Further analysis in other contexts, for example Portugal or Italy, may falsify or limit the argument by specifying additional factors, such as ideology or institutional inertia, as responsible for the centralisation/decentralisation contradiction observed in Greece.
The case highlights the insight that the crisis management literature can bring to the study of EMU. Although Perrow's (1999) contention was crafted using accidents in large technical systems, man-made disasters, such as Greece's sovereign debt crisis, are guided by similar dynamics. Public policy concerns during crises (diagnosis delay or coordination of response, for example) raise similar obstacles across substantive contexts.
The NAT argument expressed here enriches and amends EU theories of delegation. Franchino (Reference Franchino2007) points out that delegation of powers to Brussels (centralisation) is more likely when there is more conflict within the Council and higher national implementation costs. He notes that this happens under majority voting. We can amend his argument by adding the context of crisis. Even under unanimity, delegation of powers to a more centralised authority is more likely if there is conflict within the Council, high national implementation costs and a high risk of contagion. The nature of delegation depends on the creditor–debtor relationship. Because risk is shared unevenly, creditors are more likely to advocate centralisation than debtors. Being a creditor legitimises centralised intervention during a crisis to ensure effectiveness of response.
Nevertheless, greater centralisation in EU crisis management gives rise to diagnosis, coordination and compliance costs. It reduces incentives for Greek policy-makers to comply because decisions are now made by creditors; hence, blame can be shifted to them in case of failure. The findings highlight the phenomenon of responsibility drift proposed by the top-down implementation literature (Rothstein, Reference Rothstein1998). The more decision-making authority is centralised at the top, the wider the gap will be between it and the preferences of implementing agencies. The Greek case bears out the point that compliance incentives diverge widely, increasing the likelihood of implementation failure, i.e., targets will not be met. Reinforcing Schelkle's (Reference Schelkle2011: 378) argument, when creditor preferences are at odds with what the majority in parliament wants, political strife increases and all lose face.
The NAT argument also undermines the dependency on consensus conceptualised in Puetter's (Reference Puetter2012) “deliberative intergovernmentalism” during crises. True, the response of national governments aims for greater policy coherence through centralised intergovernmental coordination, but there is less dependency on consensus. On the contrary, rescue terms are increasingly dictated by creditors, leading to nested segmentation among member governments. Some governments appear to assume the role of principals, expecting others to dutifully fulfil the role of agents. EU deliberation still takes place among national governments, but not among sovereign equals.
The Greek case is not unique although it is extreme. Careful analysis distinguishes between two types of sovereign debt crises: private follies in banking or construction becoming public liabilities (for example, Ireland or Spain) and state profligacy metastasising into generalised crisis (for example, Greece or Portugal) (Stein, Reference Stein2011). The latter is far more complex than the former. The complexity of the bailout requires significant coordinating support to address the twin problems of contagion and complexity. EMU was accomplished without the stabilising effects of a central budget as a redistributive mechanism in times of crisis, effectively delaying initial attempts to prevent contagion (De Grauwe, Reference De Grauwe2006). Nevertheless, my argument predicts that growing diagnosis, coordination and compliance costs will make bailouts extremely difficult to accomplish. The paradox of country bailouts is that they require a careful calibration of urgent and decisive centralised action to address the risk of contagion and slow, deliberative and experimental governance to minimise problems posed by complexity. The EU's increasingly centralised response and its tight deadlines have predictably worked better in Ireland than in Greece. Whether Portugal suffers Greece's fate remains to be seen.
Acknowledgements
A version of this paper was presented at the annual meeting of the International Studies Association, San Diego, USA, 1–4 April 2012. I greatly benefited from comments by the participants as well as the anonymous referees and editor.
