Ex Post Evaluation of Exceptional Access Under the 2010 Stand-By Arrangement

Greece adopted an ambitious program to eliminate fiscal and current account imbalances, regain competitiveness, maintain financial stability, and boost growth and employment. While significant fiscal adjustment occurred, critical objectives such as regaining confidence and restoring growth were not achieved. The program lacked in better tailoring of IMF lending policies, avoiding undue delays in debt restructuring, attention to the political economy of adjustment, and parsimony in fiscal structural reforms. The importance of streamlining the Troika process and effective risk-sharing arrangements within the euro area has been stressed by Executive Directors.


Greece adopted an ambitious program to eliminate fiscal and current account imbalances, regain competitiveness, maintain financial stability, and boost growth and employment. While significant fiscal adjustment occurred, critical objectives such as regaining confidence and restoring growth were not achieved. The program lacked in better tailoring of IMF lending policies, avoiding undue delays in debt restructuring, attention to the political economy of adjustment, and parsimony in fiscal structural reforms. The importance of streamlining the Troika process and effective risk-sharing arrangements within the euro area has been stressed by Executive Directors.

1. An Ex-Post Evaluation (EPE) of Exceptional Access under Greece’s 2010 Stand-By Arrangement (SBA) is timely. The Fund’s rules require an EPE of exceptional access decisions within a year of the arrangement ending. This requirement is intended to ensure uniformity of treatment across members and that no large program escapes evaluation. Greece’s SBA-supported program, approved in May 2010 and cancelled in March 2012, was an exceptionally large program with access of €30 billion and warrants evaluation even though the SBA was immediately succeeded by an extended arrangement under the Extended Fund Facility (EFF). As with all EPEs, the purpose of the report is (i) to review performance against program objectives; and (ii) to evaluate program design encompassing the macroeconomic strategy, program architecture, adequacy of financing, and the case made for exceptional access.

Background to the Crisis

A. The build-up of imbalances

2. Euro accession led to an economic boom in Greece. Adoption of the euro and loose global credit conditions in the 2000s allowed Greece easy access to foreign borrowing that financed a significant expansion of government spending. Robust private credit growth following financial liberalization also served to boost household consumption. Real GDP growth averaged 4 percent from 2000–07, higher than in all euro area countries save Ireland and Luxembourg. Although asset price inflation and household indebtedness remained moderate, government debt mounted rapidly.


Long-term Government Bond Yields


Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: ECB; and Eurostat.

3. Fiscal policy was pro-cyclical. Some argue that Greece was the country that gained most from euro adoption (Fernandez-Villaverde et al, 2013) with borrowing costs falling sharply. As a direct fiscal dividend, government interest expenditure dropped from 11½ percent of GDP in the mid-1990s to 5 percent of GDP in the mid-2000s. However, these savings were more than swallowed up by increased spending on wages and pensions. The economy turned down in the wake of the Lehman crisis and the general government deficit reached 15½ percent of GDP (after incorporating data revisions), up from 4 percent of GDP in 2001. Public debt was 129 percent of GDP at end-2009, with 75 percent held by foreigners. There were also significant contingent liabilities due to public enterprises borrowing under state guarantee, while the pension system had become underfunded as a result of increasingly generous entitlements and an aging population.


General Government Operations

(Percent of GDP)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: Eurostat; and IMF staff estimates.

General Government Expenditure

(Percent of GDP)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

4. The current account deficit deteriorated. The counterpart to the decline in government saving was a sharply widening current account deficit that reached 15 percent of GDP in 2008. The sustained economic boom and a lack of competition in domestic goods and services markets kept wage and price inflation consistently above euro averages. Competitiveness, as measured by the unit labor cost (ULC) -based real effective exchange rate (REER), declined by 20–30 percent in the decade following euro adoption.


Current Account

(Percent of GDP)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: Bank of Greece; Elstat; and IMF staff calculations.

5. There were also serious doubts about the quality of Greek deficit and debt statistics. Data concerns had flared up in 2004 when upward revisions to the fiscal deficit numbers raised questions about whether Greece had ever met the Maastricht deficit criterion of an annual fiscal deficit of 3 percent of GDP. The European Commission (EC) attempted to coax Greece into reducing its deficit via the Excessive Deficit Procedure of the Stability and Growth Pact.

B. Crisis triggers

6. The global financial crisis endangered Greece’s government-led growth model. Greece’s economic boom was propelled by large foreign-funded fiscal deficits that enabled demand to outpace output. The global crisis threatened the continued financing of this growth model. After the Lehman shock in September 2008, spreads on Greek government bonds over 10-year bunds jumped to 300 bps compared to about 50 bps before the crisis. Standard and Poor’s downgraded Greece from A+ to A in January 2009 citing a loss of competitiveness worsened by the global financial crisis. Against this backdrop, Greece had become extremely vulnerable to a stop in private capital flows.

7. Data revisions unsettled markets. In October 2009, a new government took office and announced that the fiscal problem had been significantly understated. The projected budget deficit for 2009 was revised up from 4 to 12½ percent of GDP (still 3 percentage points short of the final estimate). Public debt estimates were also marked up sharply. Fitch responded by downgrading Greece’s sovereign rating from A- to BBB+. Global markets, already roiled by events in Dubai in November 2009, sold off further. Scrutiny of Greek high sovereign indebtedness was extended to other euro area countries.

C. Crisis response

8. Initially, the authorities sought a European solution. In December 2009, Greece committed to achieving fiscal consolidation via a Stability Program with the EC. At the same time, the Greek authorities ruled out the possibility of seeking Fund help. The euro area authorities seemed of like mind, indicating that financial assistance from the Fund was not “appropriate or welcome” (see Mackenzie, 2010). Nonetheless, the new government recognized that tax administration and expenditure management were extremely weak and sought technical advice from the Fund. Missions that visited Athens in early 2010 provided initial roadmaps for fiscal structural reforms.

9. Greece agreed to a fiscal consolidation plan with the EC. Greece’s 2010 Stability Program submitted to the EC in January 2010 aimed to cut the deficit from 12½ percent of GDP to 8¾ percent of GDP in 2010 and by a further 3 percentage points in 2011 and in 2012 (the “4-3-3” plan). To begin the process, the authorities announced the freezing of public sector wages, partial cancellation of civil servant bonuses, and increases in indirect taxes.

10. Markets, however, doubted whether these steps would be adequate. Financing conditions became progressively more difficult. Greece was able to raise €5 billion on two occasions in March 2010 at spreads of about 300 bps, but by the second issue, the foreign bid had dropped off appreciably. Spreads in the secondary market rose further in April, while large amortizations were coming due in May.


Sovereign Amortization Falling Due

(Billions of euros)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Source: IMF Country Report No. 09/244.

11. Against this backdrop, a request for a Fund program was made. The euro zone decided in April 2010 that the Fund should be a formal part of Greece’s rescue. Following negotiations with the newly-established Troika (the Fund, the EC and the European Central Bank (ECB)), agreement on a program to assist Greece was reached in early May. Total financing was €110 billion, of which the Fund committed €30 billion under an SBA (approved May 9). The remainder of the financing took the form of bilateral loans from euro area countries to be pooled by the EC under the Greek Loan Facility.

12. The process of building a firewall began. Contagion from Greece was a major concern for euro area members given the considerable exposure of their banks to the sovereign debt of the euro area periphery.1 To prevent the crisis spilling over, the euro area initiated the creation of a financial safety net for its member countries. In early May 2010, agreement was reached to set up the European Financial Stability Facility (EFSF) with €500 billion in financing.2 Euro area and Fund officials confirmed that these resources could be supplemented by an additional €250 billion in lending from the Fund provided on a country by country basis.


Bank Exposure to Euro Area Periphery by Country, 2009Q4

(Billions of euros)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Source: Bank for International Settlements.

13. Additional assistance came from the ECB. The ECB’s contribution was to set up the Securities Markets Program (SMP) in May 2010 that would purchase public debt securities in secondary markets. The ECB also relaxed its eligibility requirements so that Greek government debt instruments remained eligible as collateral for central bank financing despite now being below investment grade.

An Exceptional Fund Program

14. Greece’s SBA-supported program was exceptional for a number of reasons:

  • Access was the largest in Fund history. Despite representing less than 30 percent of the financing, Fund access was €30 billion, or 3,212 percent of quota. This was the largest Fund program ever relative to quota.

  • It was the first ever program with a member of the euro area. A Fund program supporting a member that had fixed its exchange rate was not unprecedented (for example, Latvia), nor was a program with a continuing member of a currency union (recent examples are St. Kitts, Benin, and Burkina Faso). Greece stood out because it is a euro area member and the euro is a reserve currency.

  • The Troika set up was novel. The Fund had recently joined forces with the EC in programs with EU members outside the euro area (Hungary, Latvia, and Romania). However, Greece was the first program in which the ECB was formally involved.

  • Public debt would remain exceptionally high during the program period. Despite a large fiscal adjustment, public debt was expected to keep rising during the program reflecting the difficult trajectory that an internal devaluation implies for debt sustainability: under a fixed exchange rate, the disinflation or deflation necessary to restore competitiveness serves to increase the debt-GDP ratio, a trend aggravated by the contraction in real GDP. This implied a public debt profile unlike other large Fund programs with the exception of Latvia which had also fixed its exchange rate. However, public debt in Latvia was appreciably lower. In Greece, debt would peak at 149 percent of GDP only in 2013. A pending data revision was expected to raise this projection by 5–7 percentage points.

  • The program required the Fund’s rules for exceptional access to be modified. To justify exceptional access, four criteria had to be met: (1) exceptional balance of payments pressures; (2) a high probability that public debt is sustainable in the medium term; (3) good prospects of regaining access to private capital markets; and (4) a reasonably strong prospect of the program’s success. The sticking point for staff was Criterion 2: even with implementation of agreed policies, uncertainties were so significant that staff was unable to vouch that public debt was sustainable with high probability. But staff favored going ahead with exceptional access because of the fear that spillovers from Greece would threaten the euro area and the global economy. A proviso was therefore added to Criterion 2 that where debt was not sustainable with high probability, a high risk of international spillover effects provided an alternative justification for exceptional access. Unusually, although this is entirely legal, this change in Fund policy was made in the context of the Greece SBA Board meeting.



(Percent of quota)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: IMF country reports; and IMF staff estimates.

Burden Sharing

(Percent of total official financing)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: IMF country reports; and IMF staff estimates.1/ Other includes €3.1 billion from EU, €1.8 billion from Nordic Countries, €0.4 billion from CZE, POL, and EST, and €0.5 billion from WB and EBRD.2/ Other includes $5 billion from U.S., $10 billion from each WB and BoJ, $4 billion from ADB, and the rest from other G7.3/ Other includes $20 billion from U.S., $10 billion from BIS and $1 billion from Bank of Canada.

Public Debt at Program Request

(Percent of GDP)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: IMF country reports; and IMF staff estimates.

Program Strategy and Implementation

15. Greece embarked on a far-reaching program of reforms in May 2010. Under the program supported by the SBA, Greece adopted an ambitious multi-year adjustment program to lower the fiscal deficit and public debt ratio, reduce domestic demand in line with supply capacity, and increase supply and competitiveness so as to invigorate investment, exports, and private sector growth. A brief history of developments under the program is provided in Box 1 which views developments under the SBA-supported program through the prism of Greece’s sovereign spreads.

16. The first pillar of the program was to drastically shrink the fiscal deficit. Fiscal adjustment was complicated by an expected 3 percentage point of GDP rise in government interest payments over 2010–14 reflecting the replacement of low-interest debt with more expensive official (and eventually private) borrowing.3 For debt to start declining, the required adjustment in the primary balance by 2014 was 14½ percentage points of GDP. To achieve this adjustment, revenue and expenditure measures amounting to 11 percent of GDP were identified that would come on top of the 5 percent of GDP in measures implemented in the first half of 2010. Reforms to tax administration and public financial management would support this effort. Major changes would also be made to the pension system to ensure its viability

17. The second pillar of the program consisted of structural reforms. Given Greece’s commitment to stay in the euro and the absence of an exchange rate lever, structural reforms were necessary to facilitate internal devaluation and boost economic growth. Specific reforms were identified to increase competition in labor and product markets, boost productivity, and improve the business environment. Positive GDP growth was projected from 2012.

18. The third pillar of the program was to preserve financial stability. The banks were viewed as vulnerable to the downturn and to an adverse feedback loop from the sovereign. The Hellenic Financial Stabilization Fund (HFSF) was set up to provide capital to banks as needed. The program included €10 billion that would be available to provide resources to the HFSF if and when such needs arose.

19. The program aimed to place Greece’s debt on a sustainable path. The combination of large fiscal adjustment and growth-enhancing structural reforms was projected to cut public debt to 120 percent of GDP by 2020. The program also envisaged a significant adjustment in the external current account, although this would be gradual since competitiveness would be regained slowly and external interest payments would rise. Program financing included €110 billion that was calculated as sufficient to cover the remaining fiscal deficit, the needs of the HFSF, and all government bonds falling due in 2010 and 2011. Market access was assumed to be regained in 2012.

20. The report is structured as follows. A review of program outcomes (page 12) shows that Greece’s SBA-supported program had some important successes, notably regarding fiscal consolidation and pension reform. Moreover, Greece was able to stay in the euro area and international systemic spillovers were relatively well-contained. However, the SBA also failed to achieve critical objectives, especially with regard to restoring growth, ensuring debt sustainability, and regaining market access. This mixed record raises questions about the design of the program (page 20) and about whether the Fund’s decision to modify its exceptional access rules to support the program was justified (page 29). The final section (page 32) reviews the possible lessons learned.

Program Outcomes

Key objectives for Greece’s SBA-supported program were to eliminate fiscal and current account imbalances, regain competitiveness, maintain financial stability, and boost growth and employment.

A. Macroeconomic outcomes

21. The economic downturn proved considerably more severe than projected. Data revisions complicate the comparison,4 but real GDP in 2012 was 17 percent lower than in 2009, compared to a 5½ percent decline projected in the SBA-supported program. The original growth projections were largely maintained until the fifth review (December 2011), but were then marked down with the expected recovery delayed until 2014. Projections for unemployment were raised in line with the severity of the contraction. The unemployment rate in 2012 was 25 percent compared to the original program projection of 15 percent. This huge rise in unemployment exceeded that which would be expected based on estimates of the Okun coefficient for Greece (see April 2010, WEO).

22. Other variables showed less divergence from projection. Inflation initially overshot the program projection due to the indirect tax increases, but subsequently came down as activity weakened. Prices fell by less than the decline in wages in part reflecting continued rigidities in product markets. Reductions in ULCs were wage driven with few indications that productivity had increased (see Box 2). The ULC REER is estimated to have declined by 9 percent during the period of the SBA. The current account deficit decreased broadly as projected, although by less than expected in cyclically-adjusted terms. There was a sharp compression of imports, but little pickup in exports.


Real GDP

(Billions of euros)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: IMF country reports; and IMF staff calculations.

Real GDP Growth

(Year-on-year percent change)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001


Greece: Evolution of Real GDP Growth Projections During 2010 and 2011


Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: Monthly Consensus Forecast and IMF Country Reports.

Current Account Balance

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: IMF country reports

Unemployment Rate


Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: IMF country reports; IMF World Economic Outlook; and IMF staff calculations.

Average CPI Inflation

(Year-on-year percent change)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

B. Fiscal policy outcomes

23. A large reduction in the fiscal deficit was achieved. The change in the primary deficit during 2010–11 was 8 percentage points of GDP, slightly above target, despite the deep recession. The authorities introduced additional measures in 2011 (Medium-Term Fiscal Strategy, amounting to 10½ percent of GDP during 2011–14) once it became clear that the initial set of fiscal measures was insufficient to deliver the consolidation target. In cyclically-adjusted terms, the primary balance improved by about 13 percent of potential GDP during the SBA.5 Estimates of the debt-stabilizing level of the primary surplus were reduced from the fifth review once private sector involvement (PSI) reduced the debt stock and lowered the average interest rate on public debt.6

24. Taxes were increased further. Revenue measures had already been taken in May 2010 under the 2010 Stability program including increases in VAT rates. Additional tax policy measures implemented during the SBA-supported program comprised increases in indirect tax rates, including further VAT rate hikes, a new property tax, and somewhat higher income taxes. Efforts were also made to strengthen tax administration and raise tax collections.

25. Expenditure measures focused on reducing public sector wages and social benefits. Public sector wages and pensions were cut through elimination of 13th and 14th monthly payments, but with safeguards intended to protect the most vulnerable. Measures implemented in 2010–11 included cuts in public sector salaries, bonuses, and allowances, and steps to reduce health care spending on drugs. Other measures included cuts in capital spending and a reorganization of subnational governments (Kalikrates). The authorities also adopted a new budget framework and put in place spending control mechanisms. Nonetheless, arrears and inadequate data reporting remained an issue throughout the SBA-supported program.


General Government Interest Expenditure

(Percent of GDP)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: IMF country reports.

Average Interest Rate on Public Debt


Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001


General Government Overall Balance

(Percent of GDP)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: IMF Country reports; and IMF staff estimates.

General Government Primary Balance

(Percent of GDP)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

General Government Operations, 2009–11

(Percent of GDP)

article image
Sources: Eurostat; and IMF staff calculations.

Percent of potential nominal GDP.

26. About half of the adjustment in the primary deficit reflected lower spending. As a result of the expenditure measures undertaken during the SBA-supported program, primary expenditure declined by 4 percentage points of GDP in 2009–11, but still exceeded the 2005 level by about the same amount. Moreover, the wage bill remained high compared with other EU countries; programs for social protection remained largely untargeted and inefficient (OECD, forthcoming); and health care reform had yet to be completed.


Changes in General Government Primary Expenditure

(Percent of GDP)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: Eurostat; and IMF staff calculations.

27. Fiscal consolidation proceeded, but momentum on fiscal structural reforms flagged. Measures taken contributed to strong upfront consolidation (wage and pension cuts, VAT hikes) and to improved long-term sustainability (pension reform). A far-reaching pension reform was approved by Parliament in July 2010 that was designed to substantially contain increases in pension benefits over the long run (see Box 3). Fund TA also led to improvements in expenditure and commitment controls and to better fiscal reporting and budgeting practices. However, tax administration reforms encountered setbacks due to political resistance and capacity constraints. Although TA led to improvements in on-time VAT filing, there were few signs by the end of the SBA-supported program that collection efficiency was being improved on a permanent basis (see Box 4).

C. Debt and financing outcomes

28. Public debt overshot program projections by a large margin. The debt sustainability analysis (DSA) in the program request included stress tests but these turned out to be mild compared to actual outcomes. At the outset of the program, debt was projected to peak at 154–156 percent of GDP in 2013 (depending on data revisions). However, by the fourth review in July 2011 (i.e., before PSI was in prospect), the end-2013 debt ratio was projected at 170 percent of GDP. As the program unfolded, the underlying debt dynamics worsened significantly because output contractions and deflation were more pronounced than expected. Lower nominal growth raised the interest rate-growth differential and led to progressively higher expected debt paths. Data revisions affecting both public debt and GDP exacerbated these trends.


Gross Public Sector Debt

(Percent of GDP)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: IMF country reports.1/ Al ready incorporated effects of debt restructuring.

Privatization Receipts

(Percent of GDP)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

29. Privatization outcomes were disappointing. The authorities signaled a strong commitment to privatization in mid-2011 with parliamentary approval of a privatization and real estate development strategy. However, despite little progress having been made to date, the fourth review made extremely optimistic assumptions about privatization with envisaged receipts rising to €50 billion compared to about €10 billion in the original program. Such receipts became less important for debt sustainability from the fifth review once PSI was in prospect, plus equity prices had by that stage come down sharply. In the fifth review, much lower projections for privatization receipts were made and in general the DSA was cast with more conservative assumptions.

30. Weaker macroeconomic outturns and eventual PSI exerted opposite effects on debt sustainability. Adjusting the debt projections under the original SBA with the macro figures used in the January 2013 review of the EFF-supported program (and for subsequent data revisions to debt and GDP) show public debt at about 180 percent of GDP in 2020. The debt restructuring, however, provided more than an offset and the latest projections show debt declining to about 120 percent of GDP by 2020.


Public Sector Debt

(Percent of GDP)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: IMF Country Reports; and IMF staff calculations.1/ As of the 2013 Article IV consultation.

31. Private creditors were able to significantly reduce their exposure. Non-resident holdings of government debt dropped sharply in 2010–12. Resident holdings of government debt initially rose, but then started to fall as well. There was a large-scale substitution from privately-held to publicly-held debt. Part of this was by design—program financing was to be used to repay maturing bonds in 2010 and 2011—but the shift was intensified by market access not being regained in 2012, as well as by SMP. Purchases of Greek government bonds under SMP created rigidities when debt was restructured as a result of the decision to exclude SMP (and euro area national central bank) bond holdings from the PSI.


General Government Debt By Owner 1/

(Billions of euros)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: Greek authorities; and IMF staff estimates.1/ Assumes that the stock of bonds in the ECB’s SMP as of end-2012 was purchased evenly over 2010-1. Bonds held by Eurosystem central banks, EC, and EIB for investment purposes are not treated as debt to official creditors.

D. Structural reforms

32. Labor market reforms encountered resistance under the SBA-supported program. Labor market reforms were initially judged to be progressing well, but a more critical view was taken in later reviews. The wage bargaining system was reformed, but there were few firm level agreements and the Fund judged labor market reforms not to have delivered enough flexibility. The absence of early actions to reduce private wages may have aggravated the job losses from the economic downturn.7

33. The product market and regulatory environment also proved resilient to change. Although the initial momentum appeared strong, reforms to product markets and the business environment failed to generate a critical mass necessary to boost growth. Program reviews reported that progress was disappointing. For example, one-stop-shops and fast-track investment approval laws were passed, but implementation was delayed. While the law liberalizing regulated professions was also passed, the deadline for requesting reinstatement of justified restrictions was postponed thus delaying implementation. This slow progress militated against realization of the productivity gains that had been hoped for in the program.

E. Financial sector

34. The banking system was perceived to be relatively sound when the program began. The bank capital ratio was 11.7 percent aided by a recapitalization in 2009, but balance sheets came under pressure from higher nonperforming loans (NPLs) once the economy weakened. Moreover, liquidity conditions tightened in 2009 due to banks losing wholesale market access and some deposit outflows.

35. As the recession intensified and liquidity tightened, the financial sector became increasingly vulnerable. Financial sector distress was a result of the protracted recession and sovereign debt problems. This was in contrast to Ireland and Spain where causality ran the other way. By 2011, deleveraging in the financial sector and restructuring of state-owned banks was perceived necessary. ATE, the largest state-owned bank and the only Greek bank to fail the Europe-wide stress tests in mid-2010, had to be recapitalized. Sizable deposit outflows began in the second half of 2011, fanned by fears of a Greek euro exit.


Funding Structure of the Banking System

(Percent of total liabilities)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: Bank of Greece; Eurostat; Haver; and Fund staff calculations.

Credit to Households

(Year-on-year percent change)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001


Eurosystem Borrowing and Emergency Liquidity Assistance

(Billions of euros)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Source: Bank of Greece.

NPLs by Category


Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

36. The ECB provided substantial and extraordinary liquidity support. As noted, from May 2010, the ECB suspended the link between sovereign credit ratings and eligibility of collateral for refinancing operations and intervened directly in the government bond market under the SMP. The ECB also began to accept uncovered bank bonds guaranteed by the government as collateral eligible for refinancing operations.

37. PSI served to eliminate the banks’ capital. Greek banks were heavily exposed to the sovereign, holding government bonds with a book value of about €40 billion (after some initial June 2011 impairments). By contrast, core capital was €22 billion, or about the same magnitude as the capital needs arising from the PSI. Only €1.5 billion was drawn from the HFSF during the SBA-supported program, but the banks’ capital needs subsequently dwarfed the HFSF provision. As of the fourth review the purpose of the HFSF changed from a means of topping up capital for banks that had tried and failed to raise private capital to providing a substantial injection of public funds for banks that had been severely affected by the PSI and the deep recession: the amount needed for the HFSF in the context of the EFF was estimated at €50 billion.

Program Design

While significant fiscal adjustment occurred during the SBA, critical objectives such as regaining confidence and restoring growth were not achieved. This raises important questions about the design of the program.

A. Should the fiscal adjustment path have been more gradual?

38. It is difficult to argue that adjustment should have been attempted more slowly. The required adjustment in the primary balance, 14½ percentage points of GDP, was an enormous adjustment with relatively few precedents,8 but was the minimum needed to bring debt down to 120 percent by 2020. Moreover, despite the starting point being slightly worse than thought to be the case when the 2010 Stability Program was drawn up,9 the SBA-supported program had already extended the period over which the Maastricht deficit target would be achieved from 3 to 5 years. Since the program only ran through mid-2013, the last part of this adjustment would occur after the program and the conditionality had ended.10 Moreover, debt would still be increasing when the program ended.

Targeted Fiscal Adjustment: Overall Deficit

(Percent of GDP)

article image
Sources: Greece’s 2010 Stability Program; and IMF Country Report No. 10/110.

39. More importantly, a flatter adjustment path would have required more than €110 billion in financing. The Greek SBA was already the highest access loan in Fund history. While the euro partners could have provided more than €80 billion in funding (although this was already more than 35 percent of Greek GDP), this would have been politically difficult. Debt restructuring could also have provided the authorities with some leeway, but as discussed below, this option was not politically feasible.

B. Should the adjustment path have been more flexible?

40. The scope for increasing flexibility was also limited. The fiscal targets became even more ambitious once the downturn exceeded expectations. In addition, the starting point moved.11 However, the automatic stabilizers were not allowed to operate and adjustments were not made to the fiscal targets until the fifth review in December 2011. While earlier adjustment of the targets could have tempered the contraction, the program would then have required additional financing. The date by which debt started to decline would also have been stretched beyond the program period.

C. Should the larger economic downturn have been expected?

41. There were a number of reasons why the actual decline in GDP was so much greater than anticipated:

  • The fiscal multipliers were too low. The question that arises is whether underestimation of the size of the fiscal multipliers in the SBA-supported program caused the depth of the recession to be underestimated. The program initially assumed a multiplier of only 0.5 despite staff’s recognition that Greece’s relatively closed economy and lack of an exchange rate tool would concentrate the fiscal shock. Recent iterations of the Greek program have assumed a multiplier of twice the size. This reflects research showing that multipliers tend to be higher when households are liquidity constrained and monetary policy cannot provide an offset (see October 2012 WEO), influences that appear not to have been fully appreciated when the SBA-supported program was designed. Aslund (2013) has also argued that there is a habitual tendency of Fund programs to be over-optimistic on growth until the economy reaches a bottom (and thereafter to underestimate the recovery).12

  • However, the deeper-than-expected contraction was not purely due to the fiscal shock. Part of the contraction in activity was not directly related to the fiscal adjustment, but rather reflected the absence of a pick-up in private sector growth due to the boost to productivity and improvements in the investment climate that the program hoped would result from structural reforms. Confidence was also badly affected by domestic social and political turmoil and talk of a Greek exit from the euro by European policy-makers.13 On the other hand, the offset to the fiscal contraction from higher private sector growth that was assumed during the program period appears to have been optimistic (see Section D below), while some of the adverse political developments were endogenous and followed from limited ownership of the program (see Section F below). A larger contraction should probably therefore have been expected, although it should be noted that market forecasters were no more accurate.14

42. In any event, a deep recession was unavoidable. Greece lost market access in the first half of 2010 with a fiscal deficit so large and amortization obligations so onerous that it is difficult to see how a severe economic contraction could have been avoided. Indeed, if Greece had defaulted, the absence of deficit financing would have required primary fiscal balance from the second half of 2010. This would have required an abrupt fiscal consolidation, and led to an evaporation of confidence and huge deposit outflow that would have most likely made the contraction in output even larger.

D. Were structural reforms sufficient to restore competitiveness and growth?

43. Structural reforms were critical to improving competitiveness. The program had to work within the constraints of the fixed exchange rate and engineer an internal devaluation. Part of the adjustment in ULCs would come from the economic slowdown that would exert downward pressure on wages. The rest would follow from structural reforms that would free up Greece’s rigid labor and product markets and raise productivity.

44. Actions were not taken to adjust private sector wages. While the program cut wages and bonuses in the public sector, there were no direct attempts to lower private sector wages. The EC took the view that forcing reductions in private wages, for example, through abolition of bonuses, was not critical: industry did not consider labor cost to be excessive and in any case exports were unlikely to be wage sensitive.15 Instead focus was on increasing the scope for wage bargaining at the firm level. The Fund agreed with the emphasis on bringing down public sector wages, noting the strong demonstration effect that this would have for the private sector.

45. A quick recovery in growth appeared optimistic. Internal devaluation was recognized to be a gradual process. In fact, the program projections implied that only about 3 percent of the estimated 20–30 percent improvement required in competitiveness would be achieved by 2013. This assumption aided the debt sustainability analysis by limiting the decline in the denominator in the debt-GDP ratio, but also raised a fundamental question about where growth would come from in the absence of an internal devaluation. The program emphasized confidence effects, regained market access, and completed structural reforms. However, even if structural reforms were transformative, a quick supply response was unlikely. Partner country growth was also expected to be weak. Nonetheless, the program assumed a V-shaped recovery from 2012.


Cumulative CPI Inflation Differential from 2011: Greece and Euro Area


Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: IMF, World Economic Outlook; and IMF staff calculations.

E. Was the mix of fiscal measures appropriate?

46. The adjustment mix seems revenue heavy given that the fiscal crisis was expenditure driven. As discussed earlier, the ballooning of the fiscal deficit in the 2000s was almost entirely due to increased expenditure. The large dose of revenue measures in the SBA-supported program can therefore be questioned, particularly since tax changes constituted almost half of the measures targeted for the first two years of the program.16 The case for indirect tax increases was that they were quick to take effect and faced less resistance than cuts in spending programs. Moreover, VAT rates were lower than the median level in Europe.

Composition of Measures

(Percent of GDP)

article image
Sources: Greek authorities; and IMF staff projections.

47. The burden of adjustment was not shared evenly across society. The public sector wage bill was cut by lowering wages and bonuses, but specific plans to downsize the number of civil servants were limited to a commitment to replace only 20 percent of those who retired. The state enterprises also remained generously staffed. By contrast, the private sector sustained enormous job losses partly because wage setting mechanisms were not liberalized. Moreover, little progress was made in checking tax evasion by high income earners. While the program recognized that it would take time to show results from improved tax administration, the absence of quick progress in collecting evaded taxes came at the cost of any demonstrable improvement in the equity of the tax burden.


Government Effectiveness, 2011 1/


Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Source: World Bank, Worldwide Governance Indicators.1/ Selected countries shown on the X axis.

F. Was there sufficient ownership and capacity for reforms?

48. Ownership of the program was limited. Staff recognized that vested interests had fiercely opposed structural reforms in Greece in the past, but were encouraged by the authorities’ strong commitment to the program. The government also had a fresh mandate from the election and a strong majority in parliament. Staff argued that difficult actions, such as cutting public sector wage levels, were being taken as prior actions, while Fund and EC technical assistance would support the program of reforms. As it turned out, ownership of the program did not extend far and little progress was made with politically difficult measures such as privatization, downsizing the public sector, and labor market reforms. There was also limited bipartisan support in parliament for the program, while relations with unions were adversarial and there was little hope that that an incomes policy could be agreed.

49. The capacity to implement reforms was overestimated. The Fund had concerns that implementation capacity might be weak based on its history of providing fiscal technical assistance to Greece. However, the extent to which administrative capacity was lacking in the public sector seems to have come as a surprise. The depth of contact afforded by annual Article IV surveillance may not have provided much insight. Moreover, World Bank indicators do not show government effectiveness in Greece to be particularly low by global standards, although they are relatively low for an advanced country.

50. Structural conditionality became very detailed. In recent years, the Fund has moved toward concentration on macro-critical structural reforms in programs and more parsimony in setting conditionality. The number of structural conditions set under the SBA-supported program was relatively large (see Box 5), although there seems to be less of a case for parsimony when formulating competitiveness-boosting reforms in a case like Greece where the exchange rate is fixed. However, the number of fiscal structural conditions was also large, and became more so as the program progressed. By the fifth review, one of the fiscal structural prior actions had nine sub-prior actions. This proliferation of conditions reflected the realization of the extent of the weaknesses in administrative capacity. While, according to available government effectiveness indicators, Greece is far from an outlier in terms of weak capacity among countries with Fund-supported programs, administrative capacities in Greece do appear to be extremely weak. Moreover, fiscal technical assistance is delivered in cooperation with the EC, whose capacity building initiatives tend to be very detailed.

G. Was the financing of the program likely to be sufficient?

51. The adequacy of the program financing required favorable assumptions. Markets were concerned about the problem of large repayment obligations in 2014 and 2015 after the program expired. The financing strategy assumed renewed market access from 2012 yet the composition of debt holders would now deter private lenders since official lenders tend to be senior creditors. Subsequent research also suggests that the market access assumption, assessed in terms of rollover rates, was sanguine compared to past experience in emerging markets facing exogenous shocks (see IMF, 2011). However, Greece’s advanced economy status and its membership of the euro area may have been considered as modifying factors.


Projected Borrowing Need Under the Program

(Billions of euros)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: IMF Country Report No. 10/110; and IMF staff estimates.

Projected Financing Sources Under the Program

(Billions of euros)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

52. The financing available for the capital needs of the banking sector was also all likely to be needed. The HFSF was established to safeguard the stability of the banking sector with up to €10 billion from program disbursements. The 2009 Article IV Consultation had reported results of banking system stress tests that showed that if a number of shocks considered were to occur simultaneously, new capital of up to €2.9 billion would be required. However, these shocks were fairly mild compared to the program projections made in May 2010, and as noted, the program assumed a quick recovery in growth that appeared optimistic. The €10 billion set aside for recapitalization under the HFSF was thus all likely to be needed.

Stress Tests of Greek Banking System

(Percent, unless otherwise indicated)

article image
Source: IMF Country Report No. 09/244

Over 2 years.

Compared to end-2008 level of 8 percent.

H. Should debt restructuring have been attempted at the outset?

53. The program was based on a number of ambitious assumptions. The preceding discussion has raised questions about whether the fiscal targets should have been less stringent and whether less optimistic projections should have been made about growth, deflation, privatization receipts, and regaining market access. Varying these assumptions would have materially affected the outlook for debt sustainability.

54. The risks were explicitly flagged. Staff made it clear that the program supported by the SBA was an ambitious program that was subject to considerable risks. The adjustment needs were huge, reforms would be socially painful, and commitment might flag. Debt was not judged to be sustainable with high probability and it would take little in the way of a deviation from program assumptions or an external shock to generate a less favorable debt trajectory. A combination of lower growth, greater deflation, higher interest rates, and larger contingent liabilities was shown to place public debt on a clearly unsustainable path (above 220 percent of GDP by 2020 and still rising). Since the shocks considered were fairly mild, this sensitivity analysis demonstrated the precariousness of the debt trajectory. For example, the deflation shock considered in the DSA (3 percent more) would not have made much difference to the internal devaluation, but would have caused debt to jump to 175 percent of GDP.


SBA: Public Debt Sustainability Analysis

(Percent of GDP)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Source: IM F Country Report No. 10/110.1/ Recognition of implicit liabilities and bank support.

55. Ex ante debt restructuring was not attempted. One way to make the debt outlook more sustainable would have been to attempt to restructure the debt from the beginning. However, PSI was not part of the original program. This was in contrast with the Fund program in Uruguay in 2002 and Jamaica in 2011 where PSI was announced upfront. In Iceland in 2008, foreign creditors (albeit private bank depositors) were bailed in ex ante via capital controls, while Hungary, Latvia, and Romania, benefitted from the Vienna Initiative that was set up to encourage foreign banks to maintain credit lines. Yet in Greece, on the eve of the program, the authorities dismissed debt restructuring as a “red herring” that was off the table for the Greek government and had not been proposed by the Fund (Papaconstantinou, 2010). In fact, debt restructuring had been considered by the parties to the negotiations but had been ruled out by the euro area. There are a number of reasons for this:

  • Some Eurozone partners emphasized moral hazard arguments against restructuring. A rescue package for Greece that incorporated debt restructuring would likely have difficulty being approved, as would be necessary, by all the euro area parliaments.

  • Debt restructuring would directly hurt the balance sheets of Greek banks. This would imply a call on the program’s financing that would exceed the amount set aside for bank recapitalization under the HFSF.

  • Debt restructuring risked contagion to other members of the Eurozone and potentially another Lehman-type event, yet the EFSF was not yet in place. European banks had large holdings of Greek bonds – but also, and of more concern given the scale of their exposure, had large holdings of the bonds of other European sovereigns that would drop in value were Greek creditors to be bailed in. For the euro zone as a whole, there might be limited gain in bailing in creditors who subsequently might themselves have to be bailed out.

56. Nonetheless, many commentators considered debt restructuring to be inevitable. With debt restructuring off the table, Greece faced two alternatives: default immediately, or move ahead as if debt restructuring could be avoided. The latter strategy was adopted, but in the event, this only served to delay debt restructuring and allowed many private creditors to escape. In fact, many observers at the time considered that the Greek program would not stave off debt rescheduling or a default. Commentators noted that the level of public debt would remain high and would be aggravated by the severe recession, while the nonconcessionary interest rates on official debt worsened the debt dynamics.17 In general, as evidenced by the behavior of bond spreads (see Box 1), markets were not convinced that the program would succeed. Statements by euro area leaders at Deauville in October 2010 that opened up the possibility of debt restructuring for Greece, contrary to its prohibition a few months earlier, also served to reinforce market doubts.

57. In the event, the SBA-supported program served as a holding operation. On the positive side, moving ahead with the Greek program gave the euro area time to build a firewall to protect other vulnerable members and averted potentially severe effects on the global economy. However, not tackling the public debt problem decisively at the outset or early in the program created uncertainty about the euro area’s capacity to resolve the crisis and likely aggravated the contraction in output.18 An upfront debt restructuring would have been better for Greece although this was not acceptable to the euro partners. A delayed debt restructuring also provided a window for private creditors to reduce exposures and shift debt into official hands. As seen earlier, this shift occurred on a significant scale and limited the bail-in of creditors when PSI eventually took place, leaving taxpayers and the official sector on the hook.

Consistency with Fund Rules and Practices

A. Was exceptional access justified?

58. The justification for exceptional access criteria and the adequacy of access levels require assessment. Exceptional access in the GRA constitutes levels beyond an annual limit of 200 percent of quota; and a cumulative limit of 600 percent of quota, net of rescheduled repurchases. Greece’s access during the first year of the SBA would amount to 1,550 percent of quota and access was expected to peak at 3,212 percent of quota in April 2013. While these levels were unprecedented, they were dwarfed by Greece’s public sector financing requirement. Moreover, the Fund’s commitment was only 30 percent of the official financing and lower access would have limited the Fund’s leverage. The exceptional access criteria apply at each review and should be assessed accordingly:

Criterion 1 (exceptional balance of payments pressures on the current or capital account) was satisfied at each stage of the program.

Criterion 2 (high probability of public debt being sustainable in the medium term) was not satisfied. As part of the summing up of the Board taken at the Greece SBA Board meeting, Criterion 2 was modified so that an exception would henceforth be made to the requirement that debt had to be judged to be sustainable with high probability in cases such as Greece where there was a high risk of international spillover effects. The systemic exception seems to have continued to be justified at each review, although the risks of contagion declined somewhat as more Greek government debt shifted to official hands.

Criterion 3 (good prospects of regaining access to private capital markets) was judged to be met. The Fund noted that successful implementation of the program and undertakings by Greece’s euro area partners to stand ready to provide additional financial assistance would help address market uncertainty and facilitate the regaining of market access. However, additional financing assurances were not strong. Key partner country authorities made some promises to corral their national banks into maintaining credit lines and debt exposures, but (as they themselves predicted) these efforts had limited effect.

Criterion 4 (a reasonably strong prospect of the program’s success taking into account institutional and political capacity to deliver adjustment) was also considered to be met. As discussed earlier, despite Greece’s mixed record of reform in the past, the Fund drew comfort from a number of factors: (i) the program was backed at the highest political levels in Greece and Europe; (ii) the most difficult actions having been taken as prior actions; and (iii) the support that TA would provide to Greece’s adjustment efforts. In the event, none of these factors proved to be compelling and ownership of the program in Greece was considerably overestimated.

The program went ahead with only Criterion 2 being modified. In the event, the judgments underlying both Criteria 3 and 4 proved to be too optimistic and with the benefit of hindsight, it is debatable whether these criteria were met at the time.

59. The program request met Fund policies on financing assurances but those weakened significantly as the program unfolded. The Fund needs to be satisfied at each review that (i) the member has secured firm financing commitments to implement the intended policies, at a minimum, for a period of 12 months and (ii) there are good prospects for full financing until the end of the program. Conditional on the program’s macroeconomic framework, the European financial commitment under the Greek Loan Facility met (i) and (ii). However, when outcomes deviated significantly from program assumptions, no new financing was committed and projected privatization receipts increased substantially, casting doubts on whether prospects for full financing until the end of the program were strong enough. 19

B. Was the SBA consistent with other Fund programs with countries in currency unions?

60. The troika arrangement was unusual. The challenge faced by the Fund in dealing with Greece was how to square its dealings with a currency union that constitutes the second largest global economic bloc while maintaining equality of treatment for all Fund members (see Pisani-Ferry et al, 2011). Greece, however, was something of a special case. The Fund has had programs with members of currency unions in the past (ECCU, CFA franc), but none where the member country was systemic. A feature of the SBA-supported program was that there was no conditionality on areas within the competency of the currency union’s central bank: however, during the course of the SBA, the ECB provided significant liquidity support to Greece and it is difficult to argue that an implicit or explicit agreement negotiated as part of the program would have required the ECB to have done more than this.

61. Nonetheless, the arrangement posed problems for program design. The Fund was required to negotiate first with the euro area countries (represented by the EC and ECB) and then with the Greek authorities. This had the advantage that the program would enjoy the necessary broader European support and that it would be consistent with euro area rules and norms. However, the Greek program was also subject to considerable uncertainty as the euro area policy response evolved. For example, the initial euro area position that debt restructuring was off the table was eventually reversed, although this took a considerable length of time. Similarly, there was an extended process before euro partners reached agreement on relaxing the fiscal stance.

C. How well did the Troika arrangement work?

62. There was no clear division of labor. Fund collaboration with the World Bank on country programs rests on an agreed division of labor. There was no such clarity in the assignment of responsibilities across the Troika. The EC needed to be involved in all aspects of the program to ensure conformity with European laws and regulations. While the Fund had experience designing fiscal adjustment, the EC had its own fiscal targets from Maastricht. The EC had structural reforms expertise, but so too did the Fund, particularly in the fiscal area. And from the Fund’s perspective, the EC, with the focus of its reforms more on compliance with EU norms than on growth impact, was not able to contribute much to identifying growth enhancing structural reforms. In the financial sector, the ECB had an obvious claim to take the lead, but was not expert in bank supervision where the Fund had specialist knowledge.

63. Areas of expertise and experience differed within the Troika. The European institutions brought an integrated view to studying the Greek economy and emphasized the extent of possible spillover effects within Europe. At least initially, this was not the perspective taken by the Fund which was more accustomed to analyzing issues with a specific country focus.20 However, the EC tended to draw up policy positions by consensus, had enjoyed limited success with implementing conditionality under the Stability and Growth Pact, and had no experience with crisis management. The Fund’s program experience and ability to move rapidly in formulating policy recommendations were skills that the European institutions lacked.

64. None of the partners seemed to view the arrangement as ideal. There were occasionally marked differences of view within the Troika, particularly with regard to the growth projections. However, the Troika in general seems to have pre-bargained positions so that differences were not on display to the authorities and did not risk slowing the program negotiations. The three institutions also have different internal procedures and the program documentation is voluminous, overlapping, and subject to varying degrees of secrecy. Nonetheless, coordination seems to have been quite good under the circumstances.

D. Should the EFF have been introduced at the beginning?

65. The three-year period of the SBA seemed relatively short. Given that the Greek program had so large a structural component, the question arises as to whether the arrangement should have been an EFF from the outset. However, while exceptional access has always been legally available under an extended arrangement, the Board as a policy matter had until the Ireland case expressed disfavor about using EFF resources as a means of providing exceptional access. In addition, the Board had generally viewed the EFF as a facility that should be used only by LICs. At the time, there was also the issue of whether the European partners were prepared to provide the longer-term financing that is a feature of EFF terms (as the Fund would not want to be the last creditor standing). Conversion of SBA into EFF was considered as early as a few months into the program but this was deferred pending consensus with the euro partners on a similar lengthening of their lending terms and until there was consensus on PSI.

Possible Lessons

66. This report concurs with the justifications that were presented for Greece’s exceptional access SBA. The Fund approved an exceptionally large loan to Greece under an SBA in May 2010 despite having considerable misgivings about Greece’s debt sustainability. The decision required the Fund to depart from its established rules on exceptional access. However, Greece came late to the Fund and the time available to negotiate the program was short. The euro partners had ruled out debt restructuring and were unwilling to provide additional financing assurances. The Fund was alert to the considerable dangers for the euro area and the global economy should Greece have been allowed to default. As the figure (adapted from the October 2012 GFSR) shows, the dislocation caused by the Lehman crisis was still recent and a renewed climate of fear in global markets had begun to emerge in May 2010. In the view of the report, this unique environment justified the provision of an exceptional access SBA to Greece.


Probability the Danish Kroner Breaks the Strong-Side of the ERM-II Boundary

(Estimated by forward rate and option markets)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: Bloomberg; and IMF staff estimates.

67. The report does not question the overall thrust of policies adopted under the SBA-supported program. Fiscal adjustment was unavoidable, as was the sharp pace of deficit reduction given that official financing was already at the limit of political feasibility and debt restructuring was initially ruled out. Structural reforms were clearly essential to restoring competitiveness. Some questions can be raised about the types of measures (overly reliant on tax increases) and structural conditionality (too detailed in the fiscal area), but the policies adopted under the program appear to have been broadly correct.

68. Nonetheless, there are possible lessons to be learned. The SBA-supported program avoided a disorderly default and limited euro-wide contagion. Greece has also been able to remain in the euro, but the recession has been deep with exceptionally high unemployment. The program did not restore growth and regain market access as it had set out to do. Major contributory factors to this lack of success were poor implementation of reform by the authorities, adverse political developments, and inconsistent policy signals by euro leaders. However, Greece’s experience also carries lessons for Fund program design and surveillance regarding the need for:

  • Better tailoring of Fund lending policies to the circumstances of monetary unions. Given the danger of contagion, there was a need to support Greece, but the Fund was worried that debt was not sustainable with high probability. In response, the exceptional access criterion was amended to lower the bar for debt sustainability in systemic cases. The baseline still showed debt to be sustainable, as is required for all Fund programs. However, the Fund emphasized the considerable risks that surrounded the baseline: growth could be lower, reforms might flag, deflation could be more severe, additional liabilities could appear. All these risks were to the downside. In the event, macro outcomes were far below the baseline and while some of this was due to exogenous factors, the baseline macro projections can be criticized for being overly optimistic.

    However, advantage could not be taken in the design of the program of one of the country’s major strengths, namely the support that Greece might ultimately receive from its euro area partners if it could demonstrate a sufficient track-record of program implementation. Adjusting lending policies to the particular circumstances of monetary unions might allow the possibility of such conditional future assistance to be recognized in program agreements. To some extent this approach is being followed in the EFF-supported program, with euro area partners indicating a willingness to provide additional financing and debt relief conditional on Greece continuing to implement the program.

    Lending policies for countries in monetary unions also need to be sensitive to the large structural component of programs when exchange rates are fixed. In this context, the Fund arrangement for Greece should ideally have been an EFF from the outset. More generally, the case for refining the Fund’s policies and framework for lending to members of currency unions needs to be explored.

  • Avoiding undue delays in debt restructuring. Upfront debt restructuring was not feasible at the outset. While the Fund began to push for PSI once the program went off track in early 2011, it took time for stakeholders to agree on a common and coherent strategy. Earlier debt restructuring could have eased the burden of adjustment on Greece and contributed to a less dramatic contraction in output. The delay provided a window for private creditors to reduce exposures and shift debt into official hands. This shift occurred on a significant scale and left the official sector on the hook.

  • More attention to the political economy of adjustment. Reform efforts in Greece under the program might have been more enduring if more visible progress had been made with regard to getting those on high incomes to pay their taxes. The program made an attempt to reflect distributional concerns by shielding those on low incomes from cuts in state pensions. The risks to public support for the program from not reducing tax evasion were also continually flagged by the Fund. The lack of political will to make clear progress with improving tax compliance was nonetheless a considerable obstacle to the program’s success.

    The Greek program also provides further evidence that the success of a program hinges centrally on the depth of its ownership. While a bipartisan agreement with the opposition was probably not possible in May 2010, the SBA-supported program attracted more bipartisan support over time including with the formation of a technocratic government in late 2011. However, significant opposition remained.

  • More parsimony in fiscal structural reforms. The Fund’s unprecedented TA programs in Greece, especially on revenue administration, may have gone beyond providing technical advice and taken on an institution building dimension. The detailed conditionality is considered macro-critical and essential given the large fiscal adjustment that is programmed and the dire need to strengthen Greek fiscal institutions. Moreover, the authorities seem to welcome the degree of Fund involvement, with no suggestion that it has become overly-intrusive. Nonetheless, these efforts are unlikely to be able to substitute for political ownership and the case for a more hands-off approach should be considered.

  • More streamlining in the Troika process. A clear division of labor of responsibilities within the Troika is difficult given the overlapping responsibilities of the institutions. There are also synergies arising from cooperation in areas with shared expertise. Nevertheless, European institutions have comparative advantage in certain structural issues that are beyond the Fund’s core areas of expertise. Options for dividing up work on areas that are not macro-critical should therefore be explored. There may also be some scope for streamlining procedures and documents to reduce the burden on the authorities.

  • More effective risk-sharing arrangements within the euro area. The Greek crisis brought to the fore shortcomings in the euro area architecture related to risk sharing and crisis response. As argued elsewhere (see IMF, 2012b), the lack of fiscal risk sharing arrangements or a true banking union within the euro area makes individual member states vulnerable to sovereign debt crises that could spill over to banking systems (or be caused by banking crises) and the real economy, without exchange rate flexibility or independent monetary policy to help with the adjustment. The euro area member states have gradually responded to the crisis in Greece and elsewhere by establishing firewalls (EFSF, ESM, and OMT) and taking steps towards a banking union, as well as strengthening fiscal rules to reduce vulnerabilities.

A Chronology of Events

  • Initial improvements in sentiment dissipated after a ratings downgrade in June 2010, but spreads tightened again with the successful completion of the first review in September.

  • However, there was an adverse market reaction to statements by euro leaders at Deauville in October 2010 that set off a public debate on the need for debt restructuring and PSI.

  • Sentiment deteriorated further in November when the estimated 2009 fiscal deficit was revised from 13½ to 15½ percent of GDP, and in December when there were general strikes and rioting in response to the labor reforms.

  • The second review of the program was completed in December 2010, but by the third review in February 2011, problems in program implementation were becoming apparent. The weak economy was making fiscal adjustment more difficult and the debt sustainability looked increasingly uncertain. Efforts were made to bring the program back on track during 2011, including ambitious privatization plans. The need for PSI was acknowledged by the fourth review in July 2011.

  • As 2011 progressed, a Greek euro exit became a serious possibility particularly after being discussed by Euro leaders at the Cannes summit in November 2011.

  • The government then announced a referendum to test the views of the Greek people. This was subsequently cancelled, but the government resigned later that month and was replaced by a technocratic government.

  • The SBA was cancelled and replaced with a new program supported by an EFF in March 2012. Sentiment subsequently turned around once the PSI came into effect. Actions by the ECB in 2012 were also extremely supportive.


Greek 10-Year Government Bond Yield, 2010–13


Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: Bloomberg; and IMF staff calculations.

Progress with Internal Devaluation under the SBA-Supported Program

Greece began the program with large price and wage competitiveness gaps due to rapid growth of labor costs and market inefficiencies. To improve competitiveness, the program initiated a comprehensive agenda of structural reforms that included reducing public sector wages; liberalizing wage-setting and loosening employment restrictions in the private sector; improving the business environment by cutting red tape; and reducing barriers to entry and market distortions in protected industries. The program began with some deregulation (fast-track investment and one-stop shop legislation), but labor market reforms were not initially deep enough to tackle entrenched labor market inflexibility.

Unit labor costs declined significantly during the program period. In real terms, the ULC-based real effective exchange rate depreciated by 9 percent from May 2010 to March 2012. The major part of the improvement in competitiveness was due to wage reductions associated with the intense recession. There were limited gains in productivity (see Kang and Shambaugh, 2013).


Gross Wages and Salaries, SA


Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: Elstat; Eurostat; and IMF staff calculations.



Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Prices have not been as flexible as wages. The CPI-based REER has a degree of upward bias as it does not discount tax increases implemented in 2010 or take into account changes in trade weights due to the redirection of exports towards emerging markets since 2009. Nevertheless, during the period of the SBA, the CPI-based real effective exchange rate depreciated by only 3 percent.


ULC Components

(Cumulative changes from 2009Q4 in logs)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: Elstat; Eurostat; and IMF staff calculations.

Labor Costs and Prices, SA


Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: Elstat; Eurostat; and IMF staff calculations.

Real Effective Exchange Rate


Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Delays in implementing service sector reforms contributed to price rigidities. Despite reform attempts, professions like pharmacology and law, as well as the transport and energy sectors, remained closed to new entrants. Continuing protection caused prices of nontradeables to remain elevated relative to the prices of tradeables until the end of 2012. In addition, the prices of tradeables compared with those of a majority of Greece’s trading partners continued to increase at or above a comparable pace through the end of 2011. Only in 2012 did the relative prices of tradeables in Greece begin to fall compared with Greece’s trading partners.

Sources: Eurostat; and IMF staff calculations.1/The EU represents nearlythree-quarters of Greece’s trading partners.

The needed reallocation of resources into exports has not been adequately supported by relative price changes. There has been limited progress with the reduction in the relative price of non-traded goods that is needed to switch production towards traded goods, while traded goods prices have also shown little decline compared with Greece’s trading partners. The ULC in non-traded goods sector has improved more than the ULC in the traded goods sector, which is a positive development, but this has not yet translated into similar improvements in relative prices due to the rigidities in product markets discussed above.

Pension Reform

Pension outlays were high before the crisis and their projected trajectory worrisome. At 16 percent of GDP, pension spending in Greece was among the highest in the EU in 2010. Key pension indicators suggested a generous system: the replacement rate (pension-to-wages) was 57 percent, the third highest in EU; the dependency rate (pensioners-to-contributors) was 29 percent, above the EU average; the official retirement age was 60, well below the OECD average of 63.2, and the average pension (including all pensions) was very close in level to that of a typical German worker with full contribution history. Pension projections pointed to a solvency problem.

Generous pension and wage increases and a worsening demographic profile were the main culprits behind the unfavorable long-term pension profile. The decade before the crisis witnessed real pension increases as well as substantial wage increases that also pushed up pensions. Numerous loopholes allowed early retirement contributing to an already high growth of pensioners. Pension spending increased by almost 5 percentage points of GDP from 2001 to 2008, outpacing other EU countries.

Other shortcomings of the system were also important. Early retirement incentives for low-income workers and high statutory replacement rates, inadequate IT and accounting systems, fragmented retirement regimes, and low compliance with contributions were some of the features of the pension system that created economic distortions and generated leakages, swelling deficits beyond sustainability.

The pension reform under the SBA-supported program tackled multiple deficiencies to bring about a sustainable long-term pension profile. Following the frontloaded pension cuts at the onset of the program, the new (PAYG) system was designed to top up the existing non-contributory social pension. The retirement age was set at 65 years and indexed to life expectancy; the base wage for the calculation of pensions became tied to lifetime wage history rather than the final five years; accrual rates were lowered, and pensions indexed to prices and GDP growth (whatever was the lowest). Other changes included harmonization of rules across funds and rationalization of the number of funds. Control mechanisms were enhanced by mandating exclusive payment of pensions through banks, and requiring periodic pension audits, monitoring of funds’ finances and periodic actuarial analysis.1

Pension System Parameters

article image
Sources: OECD; and IMF

The second stage reforms focused on deficits in supplementary and lump sum accounts and on poorly targeted social benefits. Following the pension cuts and the passage of the new law, it became clear that social spending remained high and further reforms were needed. Some of the subsequent reforms under the MTFS began tackling the still complex institutional structure, as well as the abuse of benefits, their unequal distribution, and ineffective targeting.

The pension reform was one of the main achievements of the SBA-supported program. The reform was approved in July 2010 amid widespread social discontent with austerity measures. Because the authorities wanted to accelerate the implementation, the reform was not preceded by a proper actuarial study (a structural benchmark under the program), and the public was not made fully aware of the need for reform early in the program. The reform was nonetheless essential to restoring the sustainability of the pension system —a significant achievement —although implementation risks remain. Moreover, the reform only addressed long-term structural imbalances whereas the immediate fiscal pressures on the social security system were tackled separately as part of the fiscal consolidation including by eliminating pension bonuses.

1 The pension law also included institutional reforms in the health sector providing for the financial accounting independence of health funds and integration of social security funds with the NHS.

Revenue Administration Reform during the SBA-Supported Program

Low revenue efficiency has been a chronic problem for Greece. The efficiency index of VAT collections, for example, is below average for European countries.1 Similarly, collections of personal income tax are substantially lower compared to the euro area countries, despite comparable statutory tax rates. This reflects various reasons, including institutional weaknesses of revenue administration, a large size of informal economy, a highly complex tax system with myriad preferential tax treatment and exemptions, and a large share of self-employed individuals in the business sector (OECD, 2011).


VAT C-efficiency, 2008

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: Eurostat and IMF staff calculations.

Revenue administration reform was embedded in the structural reform agenda of the program. At the beginning of the program, the authorities adopted a reform strategy that consisted of intensified anti-evasion actions against targeted taxpayers, as well as a medium-term reform plan to modernize revenue administration. In line with this, the SBA established a limited number of structural benchmarks: set up taskforces to implement the anti-evasion actions (by September 2010); table legislations to remove various obstacles for revenue administration (by February 2011); articulate a medium-term reform plan (by June 2011); and achieve quantitative performance targets (by December 2011). The design of the initial strategy and reform process throughout the SBA were supported by intensive technical assistance from the Fund, which included installment of a long-term residential advisor and peripatetic expert visits. The implementation of the benchmarks was broadly satisfactory: the first three structural benchmarks were implemented broadly as scheduled,2 and the quantitative performance targets were met except for audits of large taxpayers.

Key Performance Indicators of Tax Administration, December, 2011 1/

article image
Sources: Ministry of Finance; and IMF staff estimates.

Meeting these targets was an end-2011 structural benchmark under the 2010 SBA.

This target was reset as a prior action for the 2012 EFF and met by February, 2012.

20 percent of all assessed taxes and penalties from new audits within 3 months of assessment, and collection of 30 percent of all taxes and penalties from new audits within 6 months of assessment date.

Although program benchmarks were met, improving tax administration proved to be considerably more complex and time-consuming than expected. Looking at outcomes, the C-efficiency indicator deteriorated in 2011, and the VAT rate increases in 2010 did not bring about a higher ratio of VAT collections to GDP. While this could be due to behavioral changes of consumers (i.e., demand shifted towards goods with preferential VAT rates), it may also reflect worsened tax compliance. Indeed, the government’s anti-evasion policies do not appear to have become fully operational. For example, basic operational functions such as audits and debt collections, as well as the newly established large taxpayer unit, remained severely understaffed. The achievement of quantitative indicators did not fully reflect permanent improvements in business practices; in particular, the target for audits of wealthy taxpayers was met largely due to temporary incentives to taxpayers.


VAT Collections and Efficiency

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Sources: Eurostat; OECD; and IMF staff calculations.

The unsatisfactory progress reflects a mix of factors, including deep-rooted structural problems, worsened economic conditions, and limited ownership by the authorities. Resolving fundamental issues such as the existence of the large informal economy, the complex tax system, and the limited capacity of revenue administrators takes time. Moreover, deterioration in economic and political conditions during the crisis severely hampered anti-evasion efforts. On the other hand, the unsatisfactory progress with making reform plans functional points to limited ownership. Review by the Fund and the EC staff pointed to a lack of full political commitment to reform, absence of leadership, and continued internal resistance to reform as underlying reasons behind these developments. In hindsight, the structural benchmarks could have been more effective if the authorities had been willing to address fundamental institutional issues (e.g., increase autonomy by revamping the governance and organization structure) as well as operational aspects (e.g., ensure appropriate resource allocations to key administration functions).

Reference: OECD, 2011, “OECD Economic Surveys: Greece 2011,” OECD Publishing.

1 The VAT C-efficiency is the ratio of VAT collections to the product of the standard VAT rate and aggregate consumption. In the absence of exemptions, zero-rating (other than exports) and zero registration threshold, and no revenue leakage due to tax administration and informality factors, the C-efficiency ratio is, by construction, equal to one.2 The benchmark to table legislations to remove obstacles was completed as prior action for the 3rd review in March 2011.

Structural Benchmarks: Design and Implementation

Structural Benchmarks (SBs) focused heavily on fiscal reforms. SBs were established in the area of pensions, tax administration, public financial management, debt management, sub-national government rationalization, and government wage and employment control. Reforms were supported by extensive Fund technical assistance.


Cumulative Number of Conditions by Review

(Structural benchmarks and prior actions)

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Source: IMF staff estimates.

Competitiveness reforms flowed from the EC’s agenda. Conditionality encompassed labor markets, the transport, retail, energy, and tourism sectors, and the overall business environment. The SBA request contained only one SB related to competitiveness: the preparation of a privatization plan. The second review set an SB on reforming the collective bargaining system and the third review set one on repealing closed professions. There were numerous competiveness-related SBs in the fourth and fifth reviews. In the financial sector, establishment of the Hellenic Financial Stability Fund was an initial SB.


Average Number of Conditions per Review

Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Source: IMF staff estimates.

The total number of SBs was relatively large. The number of SBs was larger than in comparator programs (see “2011 Review of Conditionality Overview Paper”, 2012, Policy Paper and Figure 2). However, the heavy fiscal structural conditionality was a response to Greece’s weak administrative capacity, while the need for competitive-promoting reforms appears higher in a case like Greece that lacks an exchange rate tool.

Compliance with fiscal SBs was initially good, but implementation proved difficult. Up to the third review most fiscal SBs were met on time. However, the initial focus was on changes in laws and plans that were relatively easy to achieve. Limited progress was made in checking tax evasion and making the tax burden more equitable, potentially fanning public opposition to the program. There was also a growing realization that the authorities had limited capacity to implement changes on the ground and to counter resistance in the bureaucracy. The program thus began to focus increasingly on operational details, including organizational structures, audit practices, and dispute procedures that were leaving large tax debts uncollected. Performance indicators were developed to monitor progress under the tax evasion plan. By the fifth review, the content of fiscal SBs had become more detailed and encompassed multiple sub-conditions.


Share of Structural Benchmarks met in Total Due 1/


Citation: IMF Staff Country Reports 2013, 156; 10.5089/9781484387894.002.A001

Source: IMF staff estimates.1/ Cumulative number of structural benchmarks met out of structurual benchmarks due in the quarter.

Limited progress was also made with competitiveness-promoting reforms. Despite meeting formal conditionality, structural reforms also failed to be implemented due to resistance by vested interests. Growth-enhancing structural reforms that were initially cast in general terms became more specific as implementation fell short. For example, the blanket legislation to liberalize closed professions was succeeded under the EFF-supported program by detailed legislative and regulatory changes needed on a profession by profession basis.

Table 1.

Greece: Selected Economic Indicators, 2009–14

article image
Sources: National Statistical Service; Ministry of Economy and Finance; Bank of Greece; and IMF staff projections.

Based on Labor Force Survey.

Core prices exclude energy, food, alcohol, and tobacco.

Data for 2013 as of March.

Data for 2013 as of February.

Includes securitized or otherwise transferred loans from 2010 onward.

Table 2.

Greece: Quantitative Performance Criteria, 2010–13

(Billions of Euros)

article image
Source: IMF staff estimates.


Table 3.

Greece: Structural Conditionality

article image
article image
article image