Chapter 4. IMF Surveillance of the Euro Area: From Conception Through Crisis

Moisés Schwartz, and Shinji Takagi
Published Date:
June 2017
  • ShareShare
Show Summary Details
Sanjay Dhar and Shinji Takagi 


This chapter assesses the IMF’s views on the architecture of the euro area as it was being designed and its surveillance of the euro area at vital points before and during the euro area crisis. The chapter discusses IMF views on the design of the European Economic and Monetary Union (EMU) prior to its establishment, on the subsequent buildup of vulnerabilities before the crisis, and on the nature of IMF advice to respond to the crisis. Adopting such a long-term perspective enables the chapter to identify common themes in IMF surveillance viewed at from critical junctures, and to provide a broader basis for drawing lessons going forward.

The signing of the Treaty on European Union (Maastricht Treaty) in 1992 prompted a large and mostly critical academic literature on EMU. The intensity of reservations about the euro area architecture, however, dissipated following the successful launch and operation of the euro, and robust growth in the periphery. Such growth was largely driven by a surge in borrowing in the euro area periphery, sourced from euro area banks in its core, encouraged by the perceived elimination of currency and country risks. The first decade of EMU was thus characterized by credit and asset price booms in some countries, divergences in productivity growth and inflation, and widening fiscal or current account imbalances across the currency union.

Euro area economic performance in the wake of the global financial crisis, however, was weakest among the major advanced economy blocs. Output had not recovered to its pre-crisis level through 2014, while unemployment was more than double the rate in the United States, United Kingdom, or Japan, and stood at depression era levels in some countries. A number of economies experienced deflation rendering debt resolution more difficult, while low inflation for the euro area as a whole made adjustment through internal devaluation more difficult in periphery economies. This chapter discusses the advice the IMF provided to the European authorities to boost growth and to address area-wide adjustment issues when some periphery economies were in crisis.

The chapter assesses the effectiveness of IMF surveillance against this background. It is based on a review of the IMF’s published and internal documents; public statements by IMF management; and interviews with current and former euro area and member country officials, members of the IMF Executive Board, management, and staff. Where relevant, it also reflects upon the views of independent analysts, including those expressed in the academic literature. The chapter complements other chapters of this volume, which discuss the IMF’s surveillance and program engagement in individual crisis countries as well as IMF surveillance of financial and fiscal policies.

The rest of the chapter is organized as follows. The second section assesses the IMF’s views of the prospective EMU before the introduction of the euro in 1999 against the prevailing academic debate, which is elaborated upon in Annex 4.1. The third section reviews the IMF’s pre-crisis surveillance of the euro area, focusing on its discussion of the buildup of vulnerabilities before the crisis. The section also examines how effectively surveillance at the euro area level complemented country level surveillance. The fourth section assesses the IMF’s advice on policies to respond to the euro area crisis and to overcome the constraints that undermined recovery in the crisis aftermath. The final section summarizes the chapter’s key findings.

IMF Views on the Prospective EMU

Europe has neither flexible wages nor functioning labor markets, but already has mass unemployment. EMU will add to it, both on the way there and once the system is trapped in fixed rates across vastly divergent countries. (Rudiger Dornbusch, 1996)

Europe’s common market exemplifies a situation that is unfavorable to a common currency. (Milton Friedman, 1997)

Optimists about EMU think they can get along without stabilization policies other than the ECB’s commitment to price stability. They would just rely on free markets to make the necessary adjustments to economic disturbances to the union as a whole or to member states. Prices and wages will, they trust, correct fluctuations in production and employment. . . . Europe is much less well equipped than the United States to adjust to interregional disturbances to economic activity both in the strength of forces of market adjustment and in the availability of governmental fiscal responses, automatic and discretionary. (James Tobin, 1998)

The central proposal of the Maastricht Treaty signed in 1992 was that the member states of the European Community would move towards a monetary union, anchored by a single currency managed by a central bank pursuing price stability.1 There was no parallel proposal for fiscal integration or the establishment of institutions to provide a fiscal transfer mechanism within EMU. Instead, the focus of the Maastricht Treaty and the subsequent Stability and Growth Pact (SGP) of 1997 was on curtailing fiscal profligacy through rules limiting fiscal deficits and public debt, the prohibition of monetary financing of public deficits, and a no bailout clause precluding the assumption of liabilities of a member economy by the Union or its other members.2 Bank regulation and supervision remained national competencies as did deposit insurance, rather than union-level functions, and national financial systems continued to be backed by national fiscal authorities.

EMU was first and foremost a political project. Whereas the choice of exchange rate regime was a matter for each member country, the IMF’s surveillance role was to provide candid advice to members on the consistency of that choice with the member’s national policies and circumstances (IMF Articles of Agreement, Article IV). The Fund recognized the strong political backing for the prospective EMU and appropriately discussed how to improve its design. But the tone and substance of its public messages contrasted sharply with the dominant view among academic experts (Box 4.1 and Annex 4.1), as it tended to emphasize the advantages of the common currency and downplayed several of the concerns.

For example, IMF Managing Director Michel Camdessus remarked in a 1997 address:

Europe will reap a number of economic benefits from the introduction of a sound common currency. A common currency will lower transaction costs, reduce exchange risk, stimulate competition, and facilitate the broadening and deepening of European financial markets. It will also cement a larger economic space, which will then be better able to face external challenges and more impervious to adverse external shocks. All of these factors should contribute to sustained, non-inflationary growth and more rapid job creation, once structural impediments are removed.

The Maastricht monetary and fiscal framework was essentially endorsed:

The Statute of the European Central Bank guarantees the Bank’s independence and gives clear priority to low inflation in the conduct of monetary policy. Moreover, the Stability and Growth Pact, with its early warning system and various procedures for enforcing appropriate fiscal adjustment, lays out a strong framework for maintaining budgetary discipline after January 1999. At the same time, the Pact appears to leave sufficient room to tailor policy prescriptions to fit individual country circumstances. All of this augurs well for EMU and the euro.

Box 4.1.Post-Maastricht Academic Critiques of EMU

The signing of the Maastricht Treaty in 1992 led to a large and mostly critical response from some of the most prominent economists of the era. The essence of the critique was that EMU, as envisaged, did not constitute an optimum currency area (OCA)—i.e., where the benefits of a common currency in terms of lower transaction costs and elimination of exchange rate risk would outweigh the costs. Rigid wages and immobile labor within an economic region characterized by wide disparities would render adjustment to asymmetric shocks very costly in terms of output and employment. This would particularly be the case since the means to provide automatic stabilizers or discretionary fiscal policy at the union level had been ruled out by design. Under these circumstances, fiscal harmonization in terms of the convergence criteria devised for EMU membership would not suffice to ensure stability, nor could it substitute for fiscal integration, or a mechanism for fiscal transfers. By contrast, a subset of EMU nations might indeed constitute an optimum currency area.

A parallel concern related to the implementation of macroeconomic policies. A single monetary policy, likely aligned to the low inflation core, would be pro-cyclical and potentially destabilizing for EMU members with higher inflation. The fiscal rules were criticized on the grounds that they could constrain countercyclical policy, in particular, the scope for fiscal expansion in a deflationary setting. At the same time, pecuniary penalties for SGP violations would be difficult to administer in practice.

In the late 1990s, however, an alternative view emerged suggesting that closer trade links within EMU, which a single currency would promote, would lead to greater synchronization of business cycles, thereby diminishing asymmetric shocks. This view gained influence in policy circles as the final stage of EMU approached.

Distinct from OCA-related concerns, there was also concern that the Maastricht Treaty did not pay enough attention to systemic risks or the capacity to resolve them at the EMU level. In particular, fiscal stress in one country could result in systemic problems for banks and sovereigns in other countries, while a sudden stop in one country could result in sovereign default within the EMU architecture. The prospect of financial integration further pointed to the need for a central authority with a market stability mandate, and a common financial regulatory and supervisory capacity. Some experts highlighted the need for an ECB mandate for financial stability, including an explicit lender of last resort capacity.

Source: Annex 4.1.

Although the Managing Director’s remarks recognized that increased labor market flexibility would be needed to allow economies to adjust to asymmetric shocks, they were less concerned than many independent observers about the ease with which structural rigidities could be tackled. Indeed, the thrust of the advice was to avoid delays in EMU completion:

So what does all of this suggest about EMU? In my view, it suggests that this enterprise has been too long in the making, its foundation too solidly laid, and its achievement too important to European integration to court the uncertainties that would stem from its delay. In short, it is time to put to rest, once and for all, any lingering doubts about the future of EMU and to finish the job that is, in any case, so close to completion. (Camdessus, 1997)

The World Economic Outlook (WEO) which followed later that year (IMF, 1997a), analyzed these issues in greater depth. It recognized the importance of assessing countries’ ability to absorb and adjust to asymmetric shocks in light of the loss of independent monetary and exchange rate policies at the national level. It was critical of the Maastricht framework for not explicitly recognizing the importance of flexible labor markets, and emphasized the need for comprehensively addressing labor market weaknesses. And it noted that if EMU was not accompanied by further progress with structural reforms and fiscal consolidation, there were likely to be serious consequences. But in contrast to most academic observers, the WEO was considerably more sanguine about the EMU monetary and fiscal architecture, concluding that:

In the monetary and fiscal areas, the emerging policy framework appears to strike a good balance between rules and the necessary scope for the exercise of judgment in the implementation of policies. . . . Thus, the absence of a central fiscal function should not pose the problems that are commonly perceived. (Emphasis added.)

The WEO came to this conclusion by suggesting that national governments in the euro area could adjust to temporary asymmetric shocks in the context of the flexibility allowed within the SGP. Moreover, it gave prominence to the emerging view that convergence within the monetary union would be facilitated by the common currency:

. . . the incidence of asymmetric demand shocks in the EU may diminish after the euro is introduced, as asymmetric developments induced by national monetary policies or exchange rate movements within the euro area will no longer be a factor; and the discipline imposed by the Stability and Growth Pact should reduce the prevalence of fiscal shocks in the medium term.

The WEO stressed the need for both increased fiscal discipline and accelerated structural reforms to enhance labor market flexibility. But compared to outside observers it was overly optimistic that more flexible labor markets could obviate the need for greater fiscal transfers and integration. The continued emphasis on structural reforms in labor markets in the 1990s through the crisis aftermath is testament to the difficulty of achieving and sustaining reforms in this area. Moreover, the IMF’s view that the SGP would provide sufficient flexibility to respond to asymmetric shocks without a centralized fiscal capacity did not conform to the dominant academic opinion at the time.3 Instead, the IMF’s optimism on the incidence of asymmetric shocks conformed to that of European officials.

In contrast, the International Capital Markets report of November 1997 (IMF, 1997b) was more concerned about gaps in systemic risk management. Drawing on earlier work by IMF staff and others, it found ambiguity in the mechanisms for crisis resolution involving intra-European financial flows; it was critical of Maastricht’s silence on lender of last resort functions, the lack of a central authority with the mandate to ensure market stability over the euro-wide financial system, and the lack of explicit and universal cooperation and information-sharing arrangements between the European System of Central Banks (ESCB), ECB, and relevant supervisors.

Finally, in the context of considering possible use of Fund resources by a member of the euro area, an IMF Board paper (IMF, 1998) discussed a number of scenarios, which anticipated problems that would arise more than a decade later:

In principle, however, circumstances could arise in which the financial system would become segmented, such as if it were perceived that a member might depart from the union. In such circumstances, exchange risk may reappear. But even in the absence of exchange risk, lenders could still be deterred by other sources of country-specific risk, including . . . risk relating to the solvency of the national government. In the unlikely event that such risks assumed significant proportions, residents of an EMU member could find themselves unable to borrow, on suitable terms, as much as is appropriate and necessary to avoid measures destructive of national or international prosperity.

The same paper went on to predict:

Such circumstances could warrant a request for use of the Fund’s general resources by an EMU member. . . . A variety of circumstances could arise where such a need might be discerned, including through exceptional financing or liquidity support by the ECB, supplemented, perhaps, by evidence of interest rate pressures and market segmentation.

The insights from such reports were, however, lost in the euphoria of the pre-crisis period.

The IMF’s generally sanguine view was evident in its assessment of Greece’s EMU entry.4 In its last Article IV consultation with Greece (September 1999) prior to its prospective EMU entry in January 2001, the IMF raised a number of concerns. Its focus was appropriately on fiscal policy as well as structural reforms to sustain the recent reduction in inflation that Greece had attained.5 Nevertheless, the IMF’s headline statement below, as signaled in the first paragraph of the Staff Appraisal, was encouraging of a Greek EMU entry:

The stability-oriented economic policy pursued over the past years has succeeded in placing Greece squarely on the road to joining EMU by the target date of January 1, 2001—albeit with the recourse to indirect tax cuts to restrain inflation within the Maastricht criterion during the relevant reference period. The achievements to date are impressive: growth is strong, wage increases are moderate, fiscal developments are better than budgeted, and some appreciable advances have been recorded in the structural area.

The IMF’s Pre-Crisis Surveillance

A major effect of EMU is that balance of payments constraints will disappear in the way they are experienced in international relations. Private markets will finance all viable borrowers, and savings and investment balances will no longer be constraints at the national level. (European Commission, 1990)

At the launch of the euro in January 1999, the IMF adopted a dual-track approach to its surveillance of the euro area: it continued to conduct Article IV consultations, usually annually, with individual member countries of the euro area; in addition, the IMF held twice-yearly staff discussions with the institutions responsible for common policies in the euro area and issued an annual staff report followed by a formal Board meeting on the first round of discussions. The twice-yearly staff discussions would be “considered an integral part of the Article IV process for each member” (Executive Board Decision No. 11846-(98/125)).

This section discusses IMF surveillance during the first decade of the euro, focusing on how it alerted policymakers to vulnerabilities that would impact the euro area and its members. The section first provides an overview of surveillance at the euro area level (i.e., in the context of “Euro Area Policies” Article IV consultations), then discusses the effectiveness with which national and euro area level surveillance were integrated, and closes by assessing how the IMF approached the possibility of balance of payments crises in the currency union—arguably the most egregious shortcoming of its surveillance during this period.

An Overview of Surveillance at the Euro Area Level

The early years of EMU were marked by improved macroeconomic performance in several areas. The achievement of price stability and confidence in the new currency through a series of external shocks provided a welcome contrast to the turbulence European currencies had witnessed in earlier decades. Relatively strong growth in Greece, Ireland, Spain, and initially Portugal after their EMU entry appeared to vindicate the European single currency project. In this environment, the broader set of concerns among EMU critics dissipated, while attention was directed at slow growth in Germany and violation of the SGP fiscal deficit rule by Germany and France.

The IMF’s early surveillance of the euro area thus tended to be congratulatory, and applauded progress on long-sought-after structural reforms:

The launch of the euro is a defining event in the history of modern Europe. An unparalleled example of economic and political cooperation, this project has overcome many challenges in realizing the long-held dream of monetary union, initially for eleven EU countries. EMU offers participating countries and the world economy the promise of greater economic stability and enhanced economic performance. (Euro Area Policies: 1999 Article IV)

As regards structural policies, all euro-area representatives and the mission agreed that progress in product and labor market reforms was bringing about a significant transformation in the microeconomic structure of the euro area, evidenced by greater market integration and competition, falling utility prices, and increased employment content of economic growth in many countries . . . (Euro Area Policies: 2000 Article IV)

The robust performance of the periphery economies, however, was fueled by an extended surge in borrowing from banks in the euro area core, as the creation of the single currency was perceived by markets to have eliminated risk from cross-border lending, as reflected in the convergence of borrowing rates within the euro area (Figure 4.1).6 The euro area periphery’s debt to banks in the euro area core rose about eightfold between 1999 and mid-2008 (Figure 4.2). The counterpart to this borrowing was rapid credit expansion in the periphery, accompanied by higher than euro area average inflation, an appreciating real exchange rate, rising unit labor costs (Figure 4.3), and ultimately double digit current account deficit/GDP ratios (Figure 4.4). The higher inflation reduced real interest rates relative to the euro area core, further stimulating spending in the periphery. Current account deficits in the euro area periphery were almost entirely financed by private capital flows originating from within the euro area.

Figure 4.1.Ten-Year Government Bond Yields

(In percent)

Source: Eurostat.

Figure 4.2.Bank Claims of Euro Area Core to Periphery

(In billions of dollars)

Note: Claims of banks from Austria, Belgium, France, Germany, Luxembourg, and Netherlands to Greece, Ireland, Portugal, and Spain.

Source: BIS.

Figure 4.3.Unit Labor Costs Within the Euro Area

Source: Eurostat.

Figure 4.4.Current Account Balance/GDP, 1994–2014

(In percent)

Source: IMF.

The main driver of credit growth in periphery economies was growth in private sector borrowing: private debt expanded at a faster pace than public debt (Table 4.1), and was also substantially higher in absolute terms.7 In some periphery economies, local banks served as the intermediary between banks in the euro area core and their governments; as such, the figures for (gross) private debt expansion may be inflated. In the aggregate, however, public sector borrowing was not the primary factor driving higher borrowing in the euro area: between 1999 and 2007, public debt/GDP for most countries and the euro area as a whole fell (Figure 4.5); and even in Greece, the ratio was up only marginally during this period, albeit from a high initial level.8

Table 4.1.Increases in Public and Private Debt
CountryGeneral Government Debt 1998–2007 (Percentage change)Private Sector Debt 1998–2007 (Percentage change)
Source: OECD.

Greece: growth during 2001 to 2007 (i.e., since joining the euro area).

Ireland: growth during 2001 to 2007 due to lack of data dating back to 1998.

Source: OECD.

Greece: growth during 2001 to 2007 (i.e., since joining the euro area).

Ireland: growth during 2001 to 2007 due to lack of data dating back to 1998.

Figure 4.5.General Government Debt in the Euro Area

(In percent of GDP)

Note: Unweighted euro area average for the first 12 member countries.

Source: IMF, Historical Public Debt Database.

For the most part, IMF surveillance at the euro area level overlooked the fragility posed by the private credit-fueled rise in current account deficits until after the Greek crisis erupted in late 2009. It applauded the integration of the euro area bond market as signified by the convergence of yields, with little concern about the growth of private credit this was fueling:

EMU has fostered a deeper, more liquid, complete, and increasingly integrated government bond market. The convergence of yields shows the degree of integration of euro-area government bond markets. Furthermore, in the wake of EMU the volatility of yields has declined significantly and has been increasingly driven by common factors. (Euro Area Policies—Selected Issues Paper, 2005)

Monetary and credit dynamics are strong. . . . These developments, together with strong asset (notably housing) price dynamics, were seen to confirm the stimulative impact of the historically low level of interest rates. (Euro Area Policies: 2006 Article IV)

While concerns were expressed about vulnerabilities arising from rising asset prices in individual countries, they were undermined by the tendency to aggregate individual country trends, which obscured intra-euro vulnerabilities:9

At the area-wide level, accelerating credit growth does not raise prudential concerns. (Euro Area Policies: 2006 Article IV)

The tendency to downplay vulnerabilities via aggregation was also evident in the discussion of the current account, which was roughly in balance for the euro area as a whole. It was against this background that the Euro Area Policies Article IVs of 2006–07 discussed how Europe might contribute to addressing global imbalances (through structural reforms, inter alia, via greater integration of bond, equity, bank and mortgage markets), without pointing out the financing risks of the growing current account deficits in some member countries of the euro area.

Through mid-2008, when the subprime crisis in the United States had already caused significant turmoil in financial markets on both sides of the Atlantic, including the euro area, there was little warning from the IMF about reversal risks from a decade of capital flows into the euro area periphery that had fueled excessive borrowing and asset bubbles. Thus the 2008 Article IV adopted a favorable view of the stability offered by EMU and drew comfort from the improving fiscal position:

Ten years after its launch, monetary union is a distinct and promising success. EMU’s macroeconomic policy framework has brought stability. . . .

The fiscal position of the euro area improved in 2007 and now compares favorably to that in other parts of the world, marking a success for the reformed Stability and Growth Pact. In structural terms, the fiscal accounts of the euro area as a whole reached a close-to-balanced position. All excessive deficit procedures vis-à-vis euro area countries have been closed . . . Concurrently, the general government debt ratio continued to fall. (Euro Area Policies: 2008 Article IV)

In mid-2009, nine months after the Lehman collapse, the Euro Area Policies Article IV noted that the “correction of home-grown imbalances” could exacerbate the global financial turmoil. But it was only in mid-2010—after the Greek crisis had erupted—that IMF surveillance finally acknowledged the nature of systemic fragilities posed by the excessive borrowing by periphery economies since the advent of EMU:

Domestic imbalances linked to unsustainable credit and construction booms, a lack of fiscal restraint, and unsustainable wage developments all contributed significantly. As a result, unsustainable asset and demand booms emerged in some places, and a common monetary policy became increasingly ill-suited for individual parts of the region, creating destabilizing real interest and exchange rate dynamics. The exchange rate being well above fundamentals for an extended period before the crisis aggravated these dynamics: it hurt deficit countries more than surplus countries, reflecting diverging wage developments and specialization patterns. (Euro Area Policies: 2010 Article IV)

Linking National with Euro Area Level Surveillance: Fiscal and Structural Policies in Crisis Countries

Article IV consultations with member countries in the euro area highlighted to a greater extent vulnerabilities that would become relevant in the crisis, while missing others:

  • Greece. Article IV consultations with Greece following its EMU entry in 2001 consistently stressed the need for fiscal consolidation. The call for consolidation became more determined in 2004 when a data revision revealed that Greece’s fiscal deficit had never fallen below 3 percent of GDP during 1997–2003. In 2005, staff pushed for deep reforms to tax administration and expenditure management, as recommended by earlier IMF technical assistance, and the IMF continued to make the case for fiscal policy adjustment and reform to reduce the deficit throughout the pre-crisis period. The need for structural reforms, especially in labor and product markets, was also noted nearly every year as competitiveness was being eroded by rapid wage growth. In the early years, this was cast in terms of the country’s need to accelerate convergence through maintaining sustained growth; in subsequent years, the rationale was the need to contain upward price and wage pressure by improving the functioning of the labor market and promoting competition in the product market.

  • Not enough attention was paid, however, to the fundamental obstacles to enacting and sustaining the structural reforms that were assessed to be perennially needed.10 IMF staff tended to praise the authorities for any reform announced or implemented without assessing its impact; and practically any reform was cast in a positive light, albeit with a caveat that more was needed. For example, the 2007 Article IV consultation discussed in favorable terms the reforms in tax administration and expenditure control being implemented under the National Reform Program (2005–08) as well as the passage in November 2007 of the Law on Tax Evasion. In reality, very little substantive reforms were being implemented. To the contrary, the Greek government was legislating numerous structural impediments in the product market during 2004–09 (Katsoulakos, Genakos, and Houpis, 2015; Mitsopoulos and Pelagidis, 2011; Pelagidis and Mitsopoulos, 2014).

  • Ireland. Article IV consultations for Ireland took place against the background of the country’s impressive growth, accompanied by strong FDI inflows and a low stock of government debt (25 percent of GDP in 2007). While the current account remained in near balance until 2004, staff saw signs of overheating in the form of price and wage pressure, rapid private credit growth, and asset (especially house) price inflation. As a result, staff pressed for tighter fiscal policy, initially so as to make it neutral over the cycle; from 2004, the IMF began to call for modest longer-term fiscal tightening. The 2006 Article IV consultation noted that growth was too dependent on construction activity and that banks’ reliance on wholesale borrowing made the financial system vulnerable to a “change in international sentiment toward Ireland.” The 2007 Article IV noted that rapid credit growth had led to vulnerabilities, including the high share of lending to the real estate sector and the growing reliance of banks on wholesale funding, which was potentially more volatile than retail funding.

  • Even so, IMF surveillance did not discuss the risks posed by the size of the banking system relative to GDP, and the consequent high cost of resolving a banking crisis. Instead, it took comfort from the favorable indicators of banking resiliency, favorable recent stress test results, and strengthened regulation and supervision that were in line with the recommendations of the 2006 FSAP Update. Moreover, the IMF’s analysis lacked specificity. It did not pay sufficient attention to the composition of government revenue and therefore overestimated the structural fiscal position (at the peak of the housing boom, nearly one-third of government revenues were directly related to the property market). As a consequence, there was a massive discrepancy between pre- and post-crisis estimates of the cyclically adjusted balance (CAB): for example, the 2007 staff report estimated the CAB to be in small surplus in 2007, whereas the 2009 report re-estimated the 2007 CAB as a deficit of 8.7 percent (Donovan, 2017).11

  • Portugal. Article IV consultations with Portugal consistently called for tighter fiscal policy, initially in the context of emerging macroeconomic imbalances amid rapid growth. In 2000, Portugal became the first country to be subjected to the EU’s Excessive Deficit Procedure under the SGP. Staff argued that significant fiscal tightening through ambitious spending cuts was necessary to contain the wage and price pressure. As growth decelerated in 2001 and recession set in from 2002, the need for fiscal consolidation began to be framed in terms of ensuring debt sustainability against the background of population aging. Staff also highlighted the need for structural reforms in labor and product markets in order to maintain medium-term growth necessary for convergence.

  • As in other countries, however, IMF staff was too sanguine that EMU membership protected against a reversal of the capital inflows that characterized Portugal’s first decade in the euro area, when the current account deficit averaged nearly 10 percent of GDP. By mid-decade, moreover, staff appear to have misdiagnosed the root causes of Portugal’s persistent current account deficit: there was too little emphasis on rising imports and residential investment accompanied by declining private savings, and too much emphasis on declining competitiveness even though the export share of GDP and unit labor costs in manufacturing were both stable (Eichenbaum, Rebelo, and de Resende, 2017). As a result, IMF staff was insufficiently focused on the need to boost private savings and facilitate private sector deleveraging.

IMF surveillance at the euro area level ideally would have been used to warn about vulnerabilities with potentially systemic implications, or of problems afflicting more than one country. But hardly any of the issues identified as important in national surveillance were highlighted at the euro area level as a catalyst for policy adjustment or to bring to bear peer pressure. According to some European and IMF officials interviewed for this evaluation, part of the explanation may be that discussions tended to focus, perhaps justifiably, on larger, more systemically important countries, such as Germany and France.

On the fiscal policy framework, IMF staff was in favor of stronger fiscal consolidation than required under the SGP, and supported greater use of automatic fiscal stabilizers. It pointed out deficiencies such as the SGP’s focus on nominal targets that promoted pro-cyclicality, its lack of enforcement capacity, and, following the suspension of the excessive deficit procedure by the ECOFIN against France and Germany in 2003, how to restore credibility. While IMF advice was appropriate in its own right, it was unable to translate into concrete action.

Euro area surveillance called attention to the need for structural reforms, especially in the labor market, in order to raise labor market participation and to increase productivity, but policy advice was framed in such a way as to lack specific references to individual countries where action needed to take place. The advice also lacked sufficient specificity (e.g., “staff urged the authorities to make the labor market more flexible”), without giving some indication of the impact of individual measures. By not offering advice on implementation, how to overcome constraints, and the likely impact of a recommended policy measure, the likelihood of any policy action, small to begin with, was diminished further. In this context, a task force of the European Central Bank, in its review of IMF surveillance in the euro area, recommended that the IMF “provide stronger and more clearly formulated policy recommendations on structural reforms, including their estimated impact” (ECB, 2015).

IMF surveillance also did not analyze the adverse consequence of failing to address the identified fiscal or structural issues promptly, especially given the constraints posed by the monetary union. Nor did it discuss the euro area-wide implications of policy inaction in individual countries. Relevant national policies were rarely discussed in euro area consultations as a means to highlight symptoms of individual or systemic vulnerability; in particular, there was no instance in which the imbalances that were accumulating in peripheral countries were seen as posing risks and vulnerabilities for the area as a whole (Schinasi, 2012). Moreover, policy advice was not prioritized in term of macro-criticality, so that focus was not always consistent from year to year, and was overly dependent on the whim of the mission chief. Some staff interviewed for this evaluation noted that there was a limit to their ability to repeat the same advice from year to year without becoming trite.

In summary, at the risk of replicating what has already been observed elsewhere (e.g., Pisani-Ferry, Sapir and Wolff, 2011), we can characterize the manner in which the IMF’s pre-crisis surveillance fell short as a catalyst for needed policy adjustment in individual countries, in the following way. First, the IMF’s analysis or advice lacked sufficient specificity to trigger policy reaction (IMF, 2011). Second, IMF staff praised country authorities for any reform without assessing its impact. Finally, even staff’s sound advice on fiscal policy or structural reforms was not offered with sufficient urgency, as it was concurrently accompanied by the judgment that currency union membership protected against balance of payments crises (see the section “Pre-Crisis Assessment of Current Account Imbalances” below).

A common thread in these weaknesses is the IMF staff’s lack of sufficient familiarity with country-specific details. Part of the reason may be that area department missions to Greece, Ireland, and Portugal experienced more frequent turnover in mission members, as reflected in the shorter duration of country assignment, while the size of missions to these countries was also smaller (Table 4.2). These missions also tended to have less likelihood of participation from fiscal or financial experts from functional departments. Finally, the downsizing of IMF staff, which began in 2008, may have contributed to undermining surveillance in the euro area: for example, there was no Article IV consultation for Ireland in 2008 even though Irish property prices were already falling with severe repercussions for the financial sector. While this cannot be the whole story, at least it indicates the priority the IMF gave—rightly or wrongly—to less systemic countries.

Table 4.2.Staffing of Article IV Missions to Selected Euro Area Countries, 2001–08(Number of missions unless otherwise noted)
Number of MissionsAverage Size (In persons)Average Length of Country Assignment1Missions with FAD ParticipationMissions with MCM Participation
Source: IEO estimates.

Area department staff only, excluding the mission chief.

Source: IEO estimates.

Area department staff only, excluding the mission chief.

IMF surveillance of financial supervision and resolution in the early years of the monetary union served a more useful function, as discussed in more detail by Schinasi (2012) and Véron (2016). For example, the International Capital Markets report, and beginning in 2002, the Global Financial Stability Report, discussed systemic risks and vulnerabilities associated with the euro area’s financial stability architecture for preventing, managing, and resolving systemic problems. A broad concern expressed in IMF surveillance was that the nationally-oriented financial stability framework would be inadequate in handling euro-wide problems or crises. IMF staff continued to warn policymakers through the mid-2000s of the inadequacy of the prevailing system for banking crisis management and resolution and called for creating a framework with more centralized supervisory and resolution capacities. Several European officials interviewed for this evaluation expressed high marks for this aspect of the IMF’s euro area surveillance, though some IMF staff alluded to internal pressure against excessively critical analysis.12

Even so, IMF surveillance did not adequately warn about the vulnerabilities stemming from the extraordinary expansion of credit to the euro area periphery. The growing size of bank assets relative to GDP, coupled with sovereign responsibility for the rescue of national banking systems, undermined the sovereign’s own creditworthiness.13 IMF surveillance did not foresee how severely banking stress—already elevated as a result of the fallout from the global financial crisis—could become intertwined with sovereign debt problems in the euro area.14 Coupled with its failure to contemplate the possibility of a sudden stop in capital flows (discussed below), the IMF was not well positioned to provide adequate warnings on the fragility of the euro area in the face of the shocks triggered by the U.S. financial crisis.

European policymakers considered IMF surveillance “to be of little help,” with the analysis and tone “too close to the official line of the Commission and the ECB” (Pisani-Ferry, Sapir, and Wolff, 2011), a view confirmed by some European officials interviewed for this evaluation. Thus, while IMF surveillance identified many of the pertinent issues, it failed to warn the authorities and the world of the crisis that was about to engulf the currency union.

Pre-Crisis Assessment of Current Account Imbalances

A common characteristic of most Article IV reports in countries with large current account deficits was the assertion that EMU membership protected the country from external financing difficulties. In the early years of the euro this issue was nuanced in the sense that the need for (gradual) adjustment was not ruled out:

Countries that run current account positions that are fundamentally unsustainable . . . will eventually have to revise their consumption profile downwards to repay their debts, even in the context of a currency union. What is different is the nature of the required adjustment. Whereas beforehand the change in real exchange rates required to correct these imbalances could be achieved quickly through movements in the value of the currency, the adjustment will now have to come from domestic wages and prices. However, with monetary union reducing individual countries’ vulnerability to shifts in market sentiment, adjustment can probably be spread out over a longer time horizon. (Portugal: 2000 Article IV consultation)

Through 2007, the IMF continued to downplay the risk of external financing disruptions:

Given Spain’s membership in EMU and the strength of its financial sector, availability of external financing is not a constraint. (Spain: 2007 Article IV consultation)

This stance, however, appeared to soften in the 2007 Article IV on Greece, issued in 2008 (when the current account deficit was estimated at 14 percent of GDP), although even in this case, the disclaimer about external financing protection due to EMU membership was included:

In view of Greece’s EMU membership, the availability of external financing is not a concern. . . . While the risk of transmitting vulnerabilities to the euro area is very small reflecting Greece’s small relative size, large persistent current account deficits would increase the vulnerabilities to a reversal in market sentiment, leading to a corrective retrenchment of private sector balance-sheets in the face of rising indebtedness, and a possible appreciable rise in the cost of funding over time. (Greece: 2007 Article IV consultation)

IMF staff typically approached the divergent current account balances within the euro area from trade and competitiveness perspectives. For example, the 2001 Article IV consultation for Portugal suggested that, in order to contain the current account deficit, structural reforms should be implemented to shift resources from the nontradable to the tradable goods sector. Likewise in 2003, the IMF explained Greece’s large current account deficit in terms of falling competitiveness and export market share and called for wage moderation to restore competitiveness. Such analyses and recommendations were clearly pertinent. But in the process, the financing aspect of the deficit was downplayed, notwithstanding the scenario outlined by the IMF in 1998 (see the section “IMF Views on the Prospective EMU” above), in which “an EMU member could find themselves unable to borrow, on suitable terms, as much as is appropriate and necessary to avoid measures destructive of national or international prosperity.”

Staff appear to have understood the adjustment mechanism, not in terms of a sudden stop followed by a balance of payments crisis, but in terms of the price-specie-flow-like mechanism analyzed by Meade (1953) for a hypothetical common currency area in Europe.15 Such a view contrasted with an insight provided by Garber (1998, 1999), who articulated how large one-way flows of capital could result and a currency crisis ensue in EMU once the ECB’s willingness to provide unlimited credit was challenged by the market. In the event, a currency crisis of the type predicted by Garber did not materialize as the crisis countries were allowed to accumulate TARGET liabilities, that is, public capital inflows replaced the outflows of private capital (Merler and Pisani-Ferry, 2012). Even so, the mechanism of a sudden stop, and the balance of payments pressure, experienced by Greece, Ireland, and Portugal were almost exactly as predicted by Garber.

Staff’s assessment of current account deficits in the periphery appeared to reflect a standard textbook view. An influential paper by Blanchard and Giavazzi (2002), for example, argued that deficits were an expected and desirable phenomenon as euro area members became more integrated in goods and financial markets, allowing poorer countries to borrow more at lower costs to catch up with the higher income countries. Commenting on this paper, however, Gourinchas (2002) outlined how large current account deficits in Greece and Portugal could lead to illiquidity and default following a sudden stop (Annex 4.1), while Sims (2002) argued that the balance sheets and financial institutions of such countries should be carefully monitored given the prevalence of financial problems in countries experiencing large capital inflows. Sims observed that there had been a number of defaults by U.S. states in the early history of U.S. financial integration.

IMF staff’s hands-off approach to current account imbalances within the euro area contrasted with its advice elsewhere. For example, in other parts of the world, if a country was relying entirely on private capital flows to increase government consumption or fuel a real estate boom, this surely would have been a source for concern. Indeed, for several years before the global crisis, IMF staff was concerned that the United States, whose current account deficit peaked at 6 percent of GDP, could be susceptible to a dollar crisis precipitated by a drying up of capital inflows. Moreover, a number of East Asian and Latin American countries with far smaller current account deficits than in southern Europe had experienced capital account crises in the previous decade.

In fact, euro area economies were arguably more susceptible to the consequence of a change in investor sentiment compared to countries that could issue debt in their own currency.16 Individual economies experiencing credit-fueled overheating could not adjust their exchange rate, while monetary policy for the euro area as a whole could not reasonably be tailored to their needs. Their room for maneuver was therefore primarily limited to tighter fiscal and prudential policies, which proved insufficient.17

Some staff members were aware of pertinent euro area vulnerabilities. In internal comments, they were concerned by the large current account deficits in individual countries, the difficulties of designing an appropriate monetary policy across disparate parts of the euro area, and therefore of providing sufficient safeguards in countries with credit and asset booms, and by the no bail-out clause in the presence of these vulnerabilities. Even so, according to interviews with staff, there emerged an official position within the IMF, that EMU membership safeguarded individual economies from risks confronting those outside the monetary union. In considering the situation in EMU analogous to states within the United States, this position did not take into account some of the fundamental differences between the two monetary unions.18

IMF Advice During the Euro Area Crisis

A comprehensive euro area crisis management strategy was never formulated and adopted. I blame the management and staff of the IMF, the euro area countries, and other major countries for this failing. (Truman, 2013)

IMF surveillance of the euro area became more intensive in the wake of the global financial crisis. Reviews of post-crisis IMF surveillance by Pisani-Ferry, Sapir, and Wolff (2011) and ECB (2015) both indicate improvement relative to the pre-crisis period.19ECB (2015), in particular, noted that linkages and spillovers were better accounted for; there was more integration of surveillance at the bilateral and euro area levels and with multilateral exercises; and the analysis of risks had improved. And Véron (2016) found the IMF’s discussion of the need for and implementation of a banking union in the euro area was constructive.

This section addresses a different question: how effectively was IMF advice directed at containing the severity of the crisis and facilitating recovery? It is motivated by the relatively poor post-crisis economic performance of the euro area and the search for lessons for the IMF. After outlining key factors contributing to the weak recovery in the euro area, the section assesses the IMF’s advice in these areas. It finds that the IMF did not take adequate account of the euro area dimensions of the crisis, or the difficulties program countries would encounter when faced with an unconducive external environment, coupled with the constraints of adjustment in a currency union.

Economic Context: Factors Impeding Recovery

At the outset of the global financial crisis, the euro area’s public debt was comparable to that of the United States, and its overall current account was in approximate balance (versus a large U.S. deficit). Its financial sector was far smaller relative to GDP than that of the United Kingdom. Yet, the euro area economy underwent a harsher adjustment to the crisis compared to the United States, United Kingdom, and other advanced economies. Its recovery has been weaker, with the majority of euro area economies yet to recover their pre-crisis per capita output levels (Figure 4.6). Inflation stood well below target during 2013–15, rendering the intended internal devaluations in periphery economies more difficult, while unemployment in the euro area was more than double the rates of other major advanced economies in 2014, and had reached depression era levels in Greece and Spain.

Figure 4.6.Change in Real GDP Per Capita, 2007–15

(In percent)

Source: IMF, WEO, April 2016.

In the crisis aftermath, the following factors, among others, have been widely discussed as contributing to the slow recovery:

  • The euro architecture, aimed at curbing fiscal profligacy, was less suited to address the problems that confronted the euro area after the Greek crisis erupted. Not only was the scope for countercyclical fiscal policy limited, but as De Grauwe (2011) and others have pointed out, the fear of a liquidity crisis for governments lacking control over their own currency drove up bond yields across much of the euro area. Sovereign bond markets became increasingly destabilized by end-2010, with bond yields in 2011–12 reaching levels that raised questions about sovereign solvency extending well beyond Greece. This in turn aggravated distress in national banking systems given their large exposures to sovereign debt, tightening credit conditions, and setting in motion the bank-sovereign vicious cycle. It was only after ECB President Draghi’s July 2012 announcement of potentially unlimited sovereign debt purchases by the central bank (“whatever it takes”) and the ECB’s subsequent clarification of how its Outright Monetary Transactions (OMT) program would operate, did the bond market turmoil conclusively subside (Figure 4.7).20 In the meantime, however, the extended period of financial market instability was detrimental to growth, and a nascent investment recovery following the Lehman shock was reversed (Figure 4.8).

  • Macroeconomic policies in the euro area were less expansionary than in other advanced economies, reflecting asymmetric incentives to respond to the crisis, in which debtors were obliged to consolidate in the face of a hard financing constraint, while creditors faced neither incentive nor compulsion to expand. Relative to the United States, the euro area’s fiscal stimulus was weaker through 2010, and its cyclically adjusted primary balance reverted to surplus faster. This reflected not only the fiscal consolidation in periphery economies but also tightening fiscal policy in economies with more fiscal space to expand (Figure 4.9).

  • The post-crisis monetary stance in the euro area was also initially less expansionary than in other advanced economies. Interest rates were raised on two occasions in 2011 (prior to being reversed by late 2011), ECB assets declined during 2012–14 (Figure 4.10) as net asset purchases lagged and bond purchases were sterilized, and a full-fledged program of quantitative easing via sovereign bond purchases was initiated only in March 2015, six or seven years later than in the United States, United Kingdom, and Japan. Moreover, monetary conditions in program countries were considerably tighter than the euro area average even though their cyclical position called for more accommodative monetary conditions than the euro area average.

  • Weak euro area performance also stemmed from the nature of the adjustment the periphery economies were obliged to undertake. Absent the option to devalue, nominal wages in periphery economies had to decline relative to the euro area core to regain competitiveness. But the deflation or low inflation which accompanied such adjustment reduced nominal GDP and undermined debt sustainability goals. At the same time, low (and below target) inflation in the euro area undermined the effort to gain competitiveness in the periphery, prolonging their period of adjustment.

Figure 4.7.Sovereign Spreads Over 10-Year German Bund

Note: The vertical line marks the day of Mr. Draghi’s remarks in London, July 26, 2012.

Source: Eurostat, based on monthly averages.

Figure 4.8.Euro Area Gross Capital Formation

(Constant prices, 2010=100)

Source: World Bank, World Development Indicators.

Figure 4.9.Cumulated Structural Primary Fiscal Balances: 2010–2014

Note: Cumulated structural primary fiscal balances are computed as the negative y/y change in the structural balance, net of interest, as a percentage of GDP (except for Cyprus). For Cyprus, the cumulated structural primary fiscal balance is computed as the negative y/y change in the headline balance, net of interest, as a percentage of GDP.

Source: IMF, WEO, April 2016.

Figure 4.10.Central Bank Total Assets

(Percent of 2008 GDP)

Sources: Federal Reserve Bank of St. Louis and IMF.

IMF Advice on Managing the Crisis

On addressing the liquidity crisis

The IMF was an active participant in the search for institutional solutions to the euro area crisis, and pushed for more comprehensive reforms than the authorities were willing to accept (Pisani-Ferry, Sapir, and Wolff, 2011). It supported the objective of creating a credible firewall to prevent contagion, and was an advocate for expansionary monetary policy, including to support the euro area economy and financial sector. The IMF also provided useful advice on an array of financial sector policies as described in Véron (2016). But it did not offer specific advice, at least openly, on the actions needed and ultimately adopted by the ECB to resolve the crisis of confidence impacting sovereign bonds.

The IMF did not call for a re-calibration of the EMU architecture to emphasize crisis management after the Lehman Brothers bankruptcy in 2008, or as vulnerabilities in Ireland and Greece mounted in 2009. In mid-2011, the IMF emphasized that “a cohesive and cooperative approach by all euro area stakeholders” would be essential to resolve the crisis in the periphery, and supported an expansion in the role of the European Financial Stability Facility (EFSF), including, inter alia, to address the liquidity crisis:

Rapid implementation of the commitment to scale up the EFSF and a further extension of its potential uses is important to confirm that member countries “will do whatever it takes to safeguard the stability of the euro area.” (Euro Area Policies: 2011 Article IV)

Its analysis did not, however, clarify how funds drawn from the resources of euro area governments could resolve a crisis of confidence, which was already impacting the sovereign spreads of major economies such as Italy and Spain (Figure 4.7).

The IMF did not articulate the case for the ECB to play a more supportive role in addressing the liquidity crisis encompassing both sovereigns and banks through mid-2012. In this respect, the contrast between Euro Area Policies Article IV consultations during 2008–12 and 2015 is stark: the 2015 report urges the ECB to use its menu of tools for providing liquidity and purchasing assets more aggressively if needed. Yet, it was in the earlier period, when the euro area was facing a debilitating liquidity crisis, that the IMF could usefully have articulated such advice.

Advice on macroeconomic policies21

The adjustment burden facing euro area debtors could also have been eased by curtailing the deflationary bias of the post-crisis response in the euro area. This would have required some combination of more sustained fiscal expansion, more generous wage adjustments, and supplementary measures to stimulate demand in economies with fiscal space and large current account surpluses, and more expansionary monetary policy.

Fiscal policy

In the initial years of the euro area crisis, IMF staff did not analyze macroeconomic policy in individual countries from an area-wide perspective. IMF surveillance included several euro area economies in its generalized call for fiscal stimulus during 2008–09, but in mid-2010 (after the start of the troika-supported Greek program), it called for fiscal consolidation to start in each euro area economy by 2011 at the latest. In 2011, IMF staff noted:

It was agreed that maintaining easier fiscal policies in the core for the benefit of the periphery would be of little help, as direct demand effects are small. . . (Euro Area Policies: 2011 Article IV)

IMF staff supported Germany’s fiscal stimulus in 2009–10, but in 2010 it suggested that fiscal consolidation from 2011 onwards should be pursued in Germany as well to set an example for fiscal consolidation in the rest of Europe “to anchor fiscal policy in the euro area”—rather than to provide a counter to the adjustment necessitated in the euro area periphery. In support of this position, staff argued that German fiscal expansion would have limited consequences for European growth with the impact almost entirely concentrated on its immediate neighbors, fiscal policy changes in Germany would have only a small impact on the trade balance of the euro area’s peripheral economies, and would therefore be unlikely to contribute to the reduction of intra-European imbalances.

By 2014–15, however, IMF staff had changed its position on German fiscal policy, calling for it to be more expansionary. In particular, an increase in public investment was found not only to durably raise German GDP, but also to raise growth in the euro area including the periphery economies. If higher wages arose from policies that induced greater labor demand, this too would raise domestic GDP and generate beneficial regional spillovers. By this time, however, the German economy was operating closer to potential with less slack in the labor market, and hence with less grounds for engaging in fiscal expansion from a purely domestic perspective.

Monetary policy

In line with its advice to advanced economies during 2010–12 to combine fiscal consolidation with monetary expansion, IMF staff was a more consistent advocate of expansionary monetary policy during the post-crisis period. It supported the ECB’s measures to inject liquidity to the financial sector, and particularly after 2012, recognized that inadequate rather than excessive inflation would be the more pressing problem for promoting recovery in the euro area. In 2009, the staff argued for consideration of all unconventional monetary policies to support the economy and financial sector. But it did not contest—at least publicly—the ECB’s policy rate increases in 2011. From 2012 onwards, however, staff became more proactive on the need for expansionary monetary policy, stressing the risk of deflation, the need for quantitative easing, and negative interest rates.

Incorporating the euro area dimension in program design

In light of the weak economic performance of the euro area and the harsh outcomes in some of the program countries, a number of observers and interviewees have questioned the IMF’s approach to addressing the euro area crisis. For example, Truman (2013) argues that given the constraints imposed by the EMU architecture, a crisis management strategy should have been formulated at the euro area level rather than placing the entire burden of adjustment on the periphery economies, but “[the] IMF was too timid, paralyzed, or conflicted to require such steps as a condition for its participation in the Greek or subsequent programs.” He acknowledges that such a framework would have been difficult to negotiate, but notes some precedents of non-crisis countries enacting measures to support neighboring countries in crisis.22

It is beyond the scope of this chapter to discuss whether the IMF’s policy advice to euro area institutions or non-crisis euro area countries should have taken the form of conditionality, as IMF staff recently raised in its recent crisis program review (IMF, 2015).23 Regardless, it is apparent that IMF staff’s advice to the euro area during the crisis would have benefited from a more holistic approach from the standpoint of the whole euro area. Going forward, the IMF needs to ensure that policies that impact a borrowing member’s performance but are beyond the control of its authorities do not become binding constraints to recovery for members of a currency union.

Summary and Conclusions

Prior to the introduction of the euro, the IMF’s public statements were more optimistic about the benefits of a common currency and less skeptical about the design of the euro area architecture than many prominent scholars of the era. In particular, IMF surveillance noted that the lack of fiscal integration in the euro design would not pose the problems commonly perceived, and remained relatively optimistic that labor markets could be made sufficiently flexible in individual economies to compensate for the lack of an independent monetary policy. By contrast, IMF surveillance of the financial sector was highly pertinent in identifying problems with systemic risk management in the euro area as it related to the payments system and financial supervision—although it was not pursued with sufficient vigor in the pre-crisis period.

In the years prior to the euro area crisis, IMF surveillance provided little insightful discussion of the concerns that would become paramount in the crisis to follow. Vulnerabilities stemming from the extraordinary expansion of bank lending to the euro area periphery were not adequately highlighted; risks to financing double-digit current account deficits in the euro area periphery did not receive sufficient attention; there was little discussion about the risks of a sovereign debt crisis beyond calls for fiscal rectitude, or of the repercussions for banking systems from the reversal of credit and housing booms. Given these shortcomings, the intensifying nexus of banking-sovereign risks was missed. Euro Area Policies Article IV consultations, instead of focusing senior policymakers’ attention on systemic vulnerabilities, tended to downplay them, even in comparison to Article IV consultations with individual member countries.

IMF surveillance became more intensive during the crisis, with more frequent interactions at multiple levels. It made useful contributions in terms of proposals for banking and fiscal unions for EMU,24 and was a positive influence on issues of systemic reform, especially with respect to the financial sector. But it was less effective in its advice on emphasizing the indispensable role of the central bank in containing the crisis of confidence that escalated through mid-2012. Moreover, IMF surveillance did not approach the crisis from a euro area perspective: it did not appear to appreciate the constraints on recovery inherent in the adjustment prospects of the countries in crisis; and it did not call for policies in the rest of the euro area to counter the deflationary pressure and fall in demand that would emanate in the post-crisis period.

Some of these shortfalls can be traced to deficiencies in earlier episodes of IMF surveillance. Had the IMF been more skeptical about the EMU design before the euro’s introduction, it might have been more apt to raise concerns as fiscal, financial, and balance of payments risks mounted before the crisis. And had it been more concerned about such pre-crisis vulnerabilities, IMF staff may have been quicker to recognize the need to address weaknesses in the EMU’s crisis management capacity following the Lehman collapse. A common feature of IMF surveillance over almost 20 years was a tendency to side with the positions or views of authorities on most policy issues, undermining its effectiveness.

In explaining why IMF surveillance in the run-up to the financial and economic crisis failed, IEO (2011) noted two types of cognitive bias: groupthink and intellectual capture. Groupthink, the tendency among homogeneous groups to consider issues only within a certain paradigm, explains why IMF staff collectively subscribed to the view that crises were unlikely to happen in advanced economies. Likewise, IMF staff felt uncomfortable challenging the views of authorities in advanced countries, given the large number of qualified professionals working in their central banks and ministries.

These findings resonate in the case of the IMF’s role in the euro area crisis. The IMF was too deferential to official views, undermining its effectiveness as an independent technical assessor.25 This report thus confirms the findings of earlier reports. For example, Pisani-Ferry, Sapir, and Wolff (2011) concluded that the IMF had fallen “victim to a ‘Europe is different’ mindset,” and “eagerness to play a role in the complex European policy process reduced the IMF’s effectiveness.”

The desire to maintain good working relationships with authorities is both understandable and important in a membership-based institution where national authorities are the shareholders. At the same time, the Fund’s effectiveness as a guardian of global stability is diminished if it moderates the candor of its analysis for political considerations.

Recent IEO and IMF reports have provided many suggestions for strengthening surveillance and risk assessment,26 and many adjustments have already occurred. The findings of this chapter point to the need to pay special attention to the frequently observed tendency of IMF area departments to defer to the views of the authorities, especially in systemically importantly countries. Particularly in such cases, the views of departments within the IMF with expertise but less stake in their interactions with authorities could be given added weight during the review process. More generally, the IMF needs to ensure the independence of its analysis, including by fostering open discussion among staff, encouraging divergent views to be heard, and promoting greater diversity in the background and experience of key decision makers.

Annex 4.1. Post-Maastricht Academic Discussion of EMU

This annex assesses the views of prominent scholars on EMU that appear prescient with the benefit of hindsight, focusing on the period between the signing of the Maastricht Treaty and the introduction of the euro. It finds that many of them expressed serious concerns about the repercussions of EMU for member economies, and several offered useful proposals to strengthen its design.1 Although much of this discussion was guided by the optimum currency area (OCA) literature—which did not focus on banking or balance of payments risks in a monetary union—some also commented on the latter set of issues. The annex reviews each set of concerns in turn.

Optimum Currency Area Considerations

The signing of the Maastricht Treaty led to the emergence of a large academic literature on EMU. Among the critiques of EMU design were such prominent—yet ideologically diverse—economists as Dornbusch, Feldstein, Friedman, and Tobin. The tone of the early literature was generally skeptical of the long-term viability of EMU, either on optimum currency area grounds or in terms of design features, though a more supportive view later emerged emphasizing the endogeneity of OCA criteria.

As one of the early critiques of the Maastricht Treaty, Feldstein (1992) argued that EMU did not satisfy the requirements of an optimum currency area given the structure of production and trade and limited labor mobility. Without a centralized fiscal system, shocks to aggregate demand that were geographically focused, or shifts in the real equilibrium values of national exchange rates, would have large impacts on regional income and employment. Under these conditions, the loss of an independent monetary policy, in particular giving up the possibility of countercyclical monetary policy and exchange rate flexibility, would be severe, particularly in terms of employment and output, given that wages and prices adjust downwards only slowly.

Dornbusch (1996) was concerned that by abandoning the use of exchange rate adjustments, the task of adjusting for competitiveness and relative prices would fall upon the labor market. Even though Europe had vastly divergent economies, it had neither flexible wages nor flexible labor markets. As a result, the adjustment process would be frustrated, output and employment losses would dominate, and, if a region were to go into decline, deflation would have to take the place of devaluation. Dornbusch was also concerned that the costs of meeting the Maastricht criteria would be large, with minimal benefits once attained.

Likewise, Friedman (1997) argued that Europe exemplified a situation in which flexible exchange rates would be preferable to a common currency. If a monetary union was desired, however, it should follow rather than precede political unity. But a single currency introduced under unfavorable conditions would prove a barrier to the achievement of political union. He was thus concerned that the adoption of the euro “would exacerbate political tensions by converting divergent shocks that could have been readily accommodated by exchange rate changes into divisive political issues.” On the sequencing of political and monetary union, Friedman was reiterating the views not only of other economists (including some far pre-dating the Maastricht Treaty),2 but also a number of policymakers. In particular, the Bundesbank through its President expressed serious reservations about a common currency for Europe prior to the Maastricht conference (Kang and Mody, 2015).

Tobin (1998) cautioned that the combination of a monetary policy focusing solely on an inflation target, the absence of a central fiscal authority, and the prevalence of substantial structural rigidities would undermine Europe’s ability to address interregional and asymmetric shocks.3 Tobin’s skepticism of the implicit Maastricht assumption that prices and wages could be sufficiently flexible to adjust to economic shocks, made him critical of the absence of a parallel fiscal entity that would enable automatic stabilizers to operate and could administer discretionary polices at the euro area level.

Other notable critiques (from the United Kingdom) included Walters (1986, 1990) and Godley (1992). Walters argued against Britain’s membership in the exchange rate mechanism of the European Monetary System on the grounds that the loss of monetary independence would result in perverse real interest rate trends that would have destabilizing pro-cyclical effects, for example, if inflation in the United Kingdom were to exceed the European average. The so-called Walters critique was influential in subsequent debates in the United Kingdom about the merits of joining EMU.

Godley (1992) found the EMU architecture to be incomplete, and was particularly concerned about the consequences for poorer countries to survive without a mechanism for fiscal transfers: “As the treaty proposes no new institutions other than a European bank, its sponsors must suppose that nothing more is needed. But this could only be correct if modern economies were self-adjusting systems that didn’t need any management at all. . . . If a country or region has no power to devalue, and if it is not the beneficiary of a system of fiscal equalisation, then there is nothing to stop it suffering a process of cumulative and terminal decline leading, in the end, to emigration as the only alternative to poverty or starvation.”

A concern for some scholars was the particular set of problems that could arise from the proposed constellation of monetary and fiscal rules. McKinnon (1996) argued that the probability of default in EMU would rise since the option to inflate away high levels of public debt would be removed. Sims (1999) anticipated a number of problems that the euro area would later face, including: the incompatibility of the Maastricht rules in a deflationary situation where fiscal expansion may be called for; the limited effectiveness of pecuniary penalties for countries violating the Stability and Growth Pact (SGP) since fining a country in fiscal distress would be difficult and counterproductive; and the possibility that a fiscal shift in response to a speculative attack could produce multiple equilibria. De Grauwe discussed several such concerns in various editions of his textbook on European monetary integration first issued in 1992.4

Empirical work raised further questions about the susceptibility of the envisaged EMU to asymmetric shocks. For example, Bayoumi and Eichengreen (1993) found that supply shocks would be larger in magnitude and less correlated across regions in Europe than in the United States, contributing to greater difficulty in operating a monetary union in Europe—although Germany, France, the Benelux countries and Denmark would experience shocks of similar magnitude to U.S. regions.5Krugman (1993) argued that the lack of a federalized fiscal system would become more problematic in Europe over time as greater integration of markets would lead to increasingly specialized regions, which would then become more vulnerable to region-specific shocks.

It was only as the final stage of EMU approached that an alternative view emerged to argue that greater convergence would follow from the very act of introducing a common currency. In a series of papers that became influential in official circles, Frankel and Rose (e.g., 1996) provided empirical evidence suggesting that exchange rate stability promotes stronger trade links and more synchronized business cycles across countries. This evidence was widely interpreted as suggesting that the very act of introducing a common currency would facilitate satisfying OCA criteria over time.6

Banking and Balance of Payments Risks

The literature also addressed other issues, such as banking risks, the interdependency between bank exposures and sovereign risk, and sudden stops, which were not directly linked to the critiques drawing from the optimum currency area literature, but would become relevant during the crisis.

In discussing various risk scenarios Eichengreen and Wyplosz (1998) foresaw how fiscal stress in one country could impact not only the banks of that country, but lead to more systemic problems for banks and sovereigns in other countries:

The government of an EMU country gets into fiscal trouble, from which it cannot extricate itself. Investors fear suspension or (more likely) modification of payment on its public debt, and therefore sell its bonds. Its bond prices start to plummet. Banks holding those bonds find their capital impaired, inciting depositor runs. Bond markets (and indirectly banks) in other EMU countries suffer adverse repercussions, as investors in public debt become demoralized. To prevent the collapse of Europe’s banking and financial system, the ECB buys up the bonds of the government in distress.

Eichengreen and Wyplosz accordingly suggested higher capital and liquidity requirements for banks than those proposed by the Basel Accords, which were designed for countries with their own central bank, coupled with tighter limits on banks’ ability to hold sovereign bonds.

Folkerts-Landau and Garber (1994), concerned about the evolution of financial markets into “liquidity-intensive activities” and their likely convergence within the monetary union, proposed a more activist role for the ECB than envisaged in its statutes, including lender of last resort assistance, together with a parallel strengthening of the ECB’s supervisory and regulatory role. Obstfeld (1998) was also concerned about the lack of a statutory mandate for the ECB to act as a lender of last resort to rescue distressed financial institutions. And Buiter (1999) proposed that the ECB should be explicitly charged with responsibility for systemic financial stability, and the Maastricht Treaty should be revised to include the words “lender of last resort.”

Shortly after the introduction of the euro, Gourinchas (2002) challenged the prevailing official dictum by arguing that a sudden stop could lead to a situation of illiquidity, and large current account deficits in the euro area would therefore not necessarily be benign7:

Large current account deficits, even when a consequence of credible financial integration, may lead to situations of illiquidity. . . . the fact that neither Portugal’s nor Greece’s debt carries substantial spreads over that of other European countries can be taken as a sign of market confidence in these countries’ ability to honor their international obligations. But this does not mean that capital cannot or will not pull out. Even with a relatively evenly distributed maturity structure, markets could refuse to finance additional increases in debt. . . . At current levels, this would mean a sudden stop of the order of 5–7 percent of GDP. Most certainly, this would raise the specter of default. In other words, while a common currency may eliminate concerns that capital flight would force a devaluation, it does not insure against situations of illiquidity.


The terms “monetary union” and “currency union” are used interchangeably in this chapter. McCarthy (2015) argues that the euro area should not be characterized as a monetary union but rather a common currency area.

Although the central bank was prohibited from purchasing sovereign debt from the issuer, it was not prevented from acquiring sovereign debt in secondary markets. Moreover, its collateral rules facilitated commercial bank financing of public deficits with funds borrowed from national central banks. Both actions were justified by the need to buttress the conduct of open market operations and repos.

The IMF’s position on this point appears to have changed in the crisis aftermath in its discussion of the need for a fiscal union in its 2012 Euro Area Policies Article IV report. Some of the same ideas were floated earlier, for example, by the Managing Director in 2010 (Strauss-Kahn, 2010).

Many academic critiques of EMU considered a common currency suitable for a subset of prospective EMU member countries with comparable industrial structures and incomes. In effect, their reservations amounted to concerns that the EMU architecture was unsuitable if membership was to be broadened to the periphery.

Inflation was being reduced with the help of indirect tax cuts, while public debt was estimated at 106 percent of GDP in 1998.

Spreads over German sovereign borrowing rates had begun to fall in the mid-1990s in anticipation of the introduction of the euro. Krugman (2012) describes the credit fueled expansion in the periphery as the mother of asymmetric shocks, inherent to but unanticipated by the euro design. Obstfeld (2013) observes that although the growth in capital flows and banking was a global phenomenon, it was accentuated in the euro area by the integration of financial markets, and in part by a lack of credibility embodied in the no bailout clause. Others have noted that the ECB’s collateral policy, which initially did not differentiate across EMU members, also alleviated market perceptions of risk (Kopits, 2017). Sovereign yields were also compressed under EU prudential rules that sovereign debt of euro area members would carry zero risk weight for purposes of calculating regulatory capital.

For the euro area as a whole, De Grauwe (2010) reports private debt growing at an annual pace of over 35 percent during 2005–07, while public debt was growing by below 5 percent in the same period.

These trends should not be taken to infer that faster debt reduction in countries with both high and moderate public debt would not have been beneficial in managing the subsequent crisis. In particular, the decline in public debt ratio in several countries was facilitated by lower borrowing costs, which were largely spent rather than used to reduce fiscal deficits. Moreover, windfall revenues from real estate bubbles were not adequately provisioned for, as discussed below.

Article IV reports at the country level tended to discuss the risks of credit expansion more thoroughly. For example, the 2006 Article IV consultation for Spain noted that the main risks identified by the earlier Financial System Stability Assessment (FSSA) under the Financial Sector Assessment Program (FSAP) related to rapid credit growth and a potential downturn in the housing market. (It also reported that the FSSA/FSAP for Spain found a highly dynamic and competitive financial system under strong prudential supervision and regulation.)

For a succinct statement of these obstacles, see Phelps (2015).

A similar problem is evident from the difference between the IMF’s pre- and post-crisis estimates of Spain’s structural fiscal balance.

At the same time, the IMF was becoming too enamored by the U.S. approach to financial innovation as a means to embellish profitability without adequate attention to risk, encouraging, for example, Germany in 2006 to boost its bank profitability by enabling more financial innovation (IEO, 2011).

Eichengreen and Wyplosz (1998) spelled out a scenario linking the fate of sovereigns and banks even prior to the decade-long expansion in bank lending (Annex 4.1).

Exposure to U.S. subprime assets and the wholesale funding model utilized by banks were clearly not confined to the euro area. But substantial exposure to sovereigns—whose securities would be perceived as more risky—by euro area banks added an extra layer of vulnerability.

Meade (1953) argued that, for the price-specie-flow-like mechanism to work, member countries must give up domestic stabilization policies, a condition that is likely to be met only if there are union-wide stabilization policies. He also argued that, given the slow pace of price and quantity adjustment, there must be a system of providing accommodating financing: “[T]here must be some accommodating finance to cover deficits in the balances of payments during the temporary period when changed price relationships are working out their full effect” (p. 37).

Giavazzi and Spaventa (2010) noted that vulnerable euro area members were more susceptible to deteriorating investor sentiment than other weak countries around the world because they did not have their own central bank as “market maker of last resort.”

Ireland and Spain ran nominal fiscal surpluses continuously during 1999–2007, sharply reducing their public debt/GDP ratios to well below Maastricht thresholds. Ex post, their pre-crisis cyclically adjusted fiscal positions were determined to be in large deficit by the IMF—but only after the crisis. Spain’s macro-prudential policies were well regarded before the crisis.

Critical differences include: the automatic and discretionary fiscal transfers from the U.S. federal budget to the states, which comprise significant shares of state budget resources for the poorer states (averaging some 250 percent of 2009 state GDP for Mississippi, New Mexico, and West Virginia during 1990–2009 (Tressel and others, 2014); U.S. federal regulatory institutions with the authority and resources to resolve regional bank failures in contrast to the euro area; a history of state defaults to private creditors in the United States coupled with more stringent borrowing limits in the case of U.S. states, which limited their reliance on debt compared to many euro area countries; and the more highly developed and integrated U.S. capital markets.

Pisani-Ferry, Sapir, and Wolff (2011) report improvement since the start of the financial crisis in 2008 through 2010, while the focus of ECB (2015) is on how IMF surveillance changed after the 2011 Triennial Surveillance Review (IMF, 2011).

The introduction in late 2011 of the long-term refinancing operation (LTRO) of six month or longer maturity also provided significant relief to some countries embroiled in fiscal crises or credit crunches. For example, Eichenbaum, Rebelo, and de Resende (2017) indicate that Portuguese bond spreads, which continued to increase following the initial implementation of the troika-supported program for Portugal, only began to diminish following the implementation of this longer-term LTRO.

A more comprehensive discussion of IMF macroeconomic advice following the global financial crisis is contained in IEO (2014) and Dhar (2014).

For example, China was encouraged not to devalue even when its currency was under pressure during the Asian crisis. Japan exerted moral suasion with its banks to maintain its credit lines to Korean banks and extended official credit in support of private firms in Indonesia. Under the 2009 Vienna Initiative, foreign banks maintained their exposures in emerging Europe, thereby avoiding a large capital flight.

Staff’s crisis program review (IMF, 2015) clarifies that “formal conditions on union-wide policies are consistent with the Fund’s Articles of Agreement” noting that they “may be necessary for the success of the member’s program.” It also indicates that “if these approaches prove unworkable, it may be necessary to postpone Fund support until staff can give the Board an assurance that the relevant problems are being adequately addressed.”

As discussed in Véron (2016) and Kopits (2017).

The tendency of staff to defer to the views of authorities has been a well-documented weakness of IMF surveillance since the 1999 external evaluation of IMF surveillance (Crow, Arriazu, and Thygesen, 1999).

For example, see IEO (2011 and 2014) and IMF (2011 and 2014).

The annex does not aim to provide an exhaustive literature review. Jonung and Drea’s (2010) survey of the work of U.S. academic and Federal Reserve economists during 1989–2002 found a high degree of skepticism of the single currency project as designed.

For example, in discussing the European Common Market, Kaldor (1971) warned it was “a dangerous error to believe that monetary and economic union can precede a political union” because “the latter pre-supposes fiscal integration and not just fiscal harmonization.”

China’s emergence as a major importer of capital goods (to the benefit of countries such as Germany) and exporter of consumer goods (to the detriment of some euro area periphery economies) already provided an example of an asymmetric shock facing the prospective EMU.

De Grauwe (2011) later articulated how a monetary union is particularly vulnerable to changing market sentiments and multiple equilibria given the fundamental change in the nature of sovereign debt held by countries that give up the currency in which their debt is issued. In particular, investors recognize that countries with their own central banks cannot face a rollover crisis even if there is no explicit statement by the central bank to provide liquidity in a crisis.

The choice of countries suitable for a common currency varied across studies. For example, De Grauwe (1993) noted: “Whereas there is strong consensus among economists that the Twelve should not form a monetary union, there is an equally strong conviction that there is a subset of EC countries which form an optimum currency area. The minimum set of countries that could form a monetary union is generally believed to include Germany, the Benelux and possibly France.”

Referring to the influence of this work on the thinking of European policymakers, Pisani-Ferry (2013) in retrospect characterized it as a “selected reading” of the literature.

This was provided as a comment on the more sanguine view of current account deficits expressed by Blanchard and Giavazzi (2002).

    Other Resources Citing This Publication