EN 2017 Special Report NO  The Commission's intervention in the Greek financial crisis (pursuant to Article 287(4), second subparagraph, TFEU) 1977 - 2017 17 Audit team The ECA’s special reports set out the results of its audits of EU policies and programmes or management topics related to specific budgetary areas. The ECA selects and designs these audit tasks to be of maximum impact by considering the risks to performance or compliance, the level of income or spending involved, forthcoming developments and political and public interest. This report was produced by Audit Chamber IV which has a focus in the areas of regulation of markets and competitive economy. Mr Baudilio Tomé Muguruza, Dean of this Chamber, is the reporting Member. He was supported in the preparation of the report by Daniel Costa de Magalhães; Ignacio García de Parada Miranda; and Simon Dennett from his private office; Zacharias Kolias, director; and Kamila Lepkowska, head of task. The audit team consisted of Efstathios Efstathiou, Athanasios Koustoulidis, Adrian Savin, Giuseppe Diana, Marion Schiefele and Natalia Krzempek. From left to right: Daniel Costa de Magalhães, Giuseppe Diana, Marion Schiefele, Adrian Savin, Kamila Lepkowska, Simon Dennett, Baudilio Tomé Muguruza, Ignacio García de Parada Miranda, Efstathios Efstathiou, Zacharias Kolias, Natalia Krzempek. 2 CONTENTS Paragraph Abbreviations Executive summary About this report About the Greek Adjustment Programmes How we conducted our audit What we found What we recommend Introduction The three Greek Economic Adjustment Programmes Programme objectives The programme partners and the role of the European institutions The programming process Audit approach Observations Part I – Management of the economic adjustment programmes for Greece Arrangements for designing and monitoring the programmes’ conditions I - XIV I II - V VI - VII VIII - XIII XIV 1 - 15 1-6 7 8 - 10 11 - 15 16 - 20 21 - 130 21 - 54 22 - 40 Urgency at the inception of the programmes 22 - 23 The design procedures and outcomes 24 - 28 Overall growth strategy 29 - 30 Political instability during the programme period 31 - 32 Monitoring procedures 33 - 35 Monitoring practices 36 - 40 Cooperation with programme partners 41 - 43 The programmes’ economic assumptions 44 - 54 3 The methodological strengths and weaknesses 45 - 48 Sensitivity tests and justification of assumptions 49 - 51 Accuracy and adjustments 52 - 54 Part II - Design and implementation of the reforms Taxation 55 - 102 58 - 66 Conditionality design - tax policies 59 - 62 Conditionality design - tax enforcement 63 - 64 Implementation and results 65 - 66 Public administration reform 67 - 76 Conditionality design 68 - 73 Implementation and results 74 - 76 Financial-sector reforms Conditionality design Expected financial results Labour market reforms Conditionality design Implementation and results Part III - Achievement of programme objectives Ensuring fiscal sustainability Access to markets 77 - 94 79 - 93 94 95 - 102 97 - 100 101 - 102 103 - 130 107 - 116 108 Use of funds and post-programme financing 109 - 110 Fiscal consolidation 111 - 114 Sustainability of the pension system 115 - 116 Restoring financial stability Solvency Credit risk and asset quality 117 - 123 118 119 - 120 4 Liquidity 121 Profitability 122 - 123 Reviving growth 124 - 130 Economic growth and its context 125 - 128 Prices and unemployment 129 - 130 Conclusions and recommendations 131 - 146 Management of the Economic Adjustment Programmes for Greece 131 - 138 Design and implementation of the reforms 139 - 143 Achievement of programme objectives 144 - 146 Replies of the Commission Annex I – The chronology of the Greek economic crisis Annex II – Capital controls Annex III - Weaknesses in compliance monitoring Annex IV – Taxation reforms Annex V - Evolution of the funding gap methodology Annex VI - Accuracy of the macroeconomic projections Annex VII - Delayed and non-complied conditions Annex VIII – Public administration Annex IX - Financial sector reforms Annex X – Financial stability Annex XI – Sustainable growth 5 ABBREVIATIONS AMC: Asset Management Company AROP: ‘Administrative Reform’ Operational Programme, for the programming period 20072013 BoG: Bank of Greece DG ECFIN: European Commission Directorate-General for Economic and Financial Affairs EAP: Economic Adjustment Programme EC: European Commission ECB: European Central Bank ECJ: European Court of Justice EDP: Excessive Deficit Procedure EEFs: European Economic Forecasts EFSF: European Financial Stability Facility EFSM: European Financial Stabilisation Mechanism ENFIA tax: Uniform tax on real estate property ESA: European System of Accounts EPL: Employment protection legislation ESM: European Stability Mechanism FDI: Foreign Direct Investment FG: Funding Gap GDP: Gross Domestic Product 6 GLF: Greek loan facility HFSF: Hellenic Financial Stability Fund IMF: International Monetary Fund LFA: Loan facility agreement MEFP: Memorandum of Economic and Financial Policy MoU: Memorandum of Understanding NPL: Non-Performing Loans OECD: Organisation of Economic Co-operation and Development PA: Public Administration PPC levy: Public Power Corporation levy PSC: Point of Single Contact PSI: Private Sector Involvement SGPR: General Secretariat for Public Revenue SSM: Single Supervisory Mechanism T-bills: Short-term debt instruments issues for a term of one year or less TFEU: Treaty on the Functioning of the European Union TFGR: Task Force for Greece TMoU: Technical Memorandum of Understanding ULCs: Unit labour costs 7 EXECUTIVE SUMMARY About this report I. We examined the European Commission’s management of the three Economic Adjustment Programmes for Greece, bearing in mind the institutional set-up of the different financial assistance instruments used. In relation to the ongoing programme, the audit focused only on the design aspects. Funding for the first programme (GLF), in 2010, was 110 billion euros; for the second (EFSF; 2012) it was 172.6 billion euros and for the third (ESM; 2015) it was 86 billion euros. As of mid-2017, Greece still requires external financial support and we found that the objectives of the programmes were met only to a limited extent. Overall, the programmes’ design did make the progress of reform in Greece possible, but we found weaknesses. We make a number of recommendations to the Commission for future support programmes. About the Greek Adjustment Programmes II. From the time of its entry into the euro, Greece benefitted from an economic boom fuelled by easy access to borrowing and generous fiscal policy. But the 2008-2009 global financial crisis exposed the country’s vulnerabilities: growing macroeconomic imbalances, large stocks of public and external debt, weak external competitiveness, an unsustainable pension system and weak institutions. These combined with revelations about misreporting of official statistics impacted international confidence. The price which Greece had to pay to borrow on the financial markets became unsustainable and in April 2010 the country requested financial assistance from the Euro area member states and the IMF. III. The first Economic Adjustment Programme for Greece was agreed in 2010 and included funding of 110 billion euros. But despite the fiscal and structural measures undertaken and the debt restructuring in 2012, it was not sufficient for the country to return to market financing. Two further programmes were therefore agreed, for 172.6 billion euros in 2012 and 86 billion in 2015. IV. The Adjustment Programmes aimed to address the economic imbalances in Greece and thereby to prevent contagion between the Greek economic crisis and the rest of the 8 euro area. They sought to establish a sound and sustainable economic and financial situation in Greece and to restore its capacity to finance itself fully on the financial markets. Assistance was subject to policy conditions, defined by agreement between the Greek authorities and the lenders. The conditions covered virtually all the functions of the Greek state and addressed three main objectives: fiscal sustainability, financial stability and restoration of growth. The European Commission verified and reported on Greece’s compliance with the conditions. V. The Commission managed the Adjustment Programmes in liaison with the European Central Bank on behalf of the European lenders for each programme: the Euro area member states for the first programme, the European Financial Stability Facility for the second and the European Stability Mechanism for the third. How we conducted our audit VI. We examined whether the Commission’s management of the Adjustment Programmes for Greece was appropriate. In particular, we asked: • Did the Commission have appropriate arrangements in place for managing the programmes? VII. • Were the policy conditions appropriately designed and effectively implemented? • Did the adjustment programmes meet their main objectives? In line with the ECA’s mandate to audit the operational efficiency of the management of the ECB, we have attempted to examine the Bank’s involvement in the Greek Economic Adjustment Programmes. However, the ECB questioned the Court’s mandate in this respect; did not provide sufficient amount of evidence and thus we were unable to report on the role of the ECB in the Greek programmes. What we found VIII. At the start of the Greek programme, the Commission had no experience in the management of such a process. Procedures were established after almost a year, but they focused on the formal arrangements for approval of documents, information flows and the timeline for disbursements. There were no specific internal Commission guidelines on the 9 actual design of the programmes’ conditions, for example in terms of scope or level of detail. Despite a growing number of conditions, the first and second programmes did not adequately prioritise their relative importance and they were not embedded in a broader strategy for the country. The Commission did put in place a functioning system for assessing the conditions, but we found consequential weaknesses, in particular for the assessment of implementation of structural reforms. IX. Despite the complex institutional arrangements within the programmes, the operational details of the Commission’s co-operation with programmes partners, primarily the IMF and the ECB, were never formalised. X. The overall economic framework for the design of programmes was made up of the funding gap calculations and macroeconomic projections. The Commission regularly updated its analysis in this respect and the accuracy of projections was similar when compared to other international organisations. However, we found weaknesses in documentation, justification of the assumptions and quality controls. XI. An in-depth analysis of the design and implementation of reforms in four crucial policy fields (taxation, public administration, labour market and the financial sector) provides a mixed picture. The taxation and public administration reforms brought fiscal savings, but the implementation of structural components was weaker. The labour market has become more flexible and competitive, while further regulatory changes are still on the way in the third programme. The financial sector was substantially re-structured, but this came at a cost of over 45 billion euros injected into the banking system, out of which only a small part can potentially be recovered. Across all policy fields, the implementation of a number of key reforms happened with significant delays or was not effective. XII. Overall, the design of the conditions did make the progress of reform possible, but we found weaknesses. Some key measures were not sufficiently justified or adapted to specific sector weaknesses. For others, the Commission did not comprehensively consider Greece’s implementation capacity in the design process and thus did not adapt the scope and timing accordingly. We also found cases of conditions with too narrow a scope to address key 10 sector imbalances and late inclusion of measures addressing key imbalances in the programme. XIII. The Commission did not carry out a comprehensive evaluation of the first two programmes, although such an analysis could be pertinent for adjusting the reform process. As of mid-2017, Greece still requires external financial support, which indicates that the earlier programmes, also due to implementation weaknesses, were not able to restore the country’s ability to finance its needs on the markets. The specific objectives of the programmes were met only to a limited extent: • Return to growth: GDP shrank during the programmes by over a quarter and Greece did not return to growth, as initially envisaged, in 2012. • Fiscal sustainability: there was large scale fiscal consolidation in terms of structural balances. But due to adverse macroeconomic developments and interest costs on existing debt, the debt to GDP ratio kept increasing. • Financial stability: the programmes ensured short-term financial stability, but were unable to avert a sharp deterioration of the banks’ balance sheets, primarily due to adverse macroeconomic and political developments, and the banks’ ability to provide finance to the real economy was restricted. What we recommend XIV. The European Commission should: (a) Improve the procedures for the design of support programmes, in particular by outlining the scope of any analytical work necessary to justify the content of the conditions. (b) Better prioritise the conditions and specify measures urgently needed to address the imbalances that are crucial for achieving the objectives of the programmes. (c) Where relevant to address the underlying economic imbalances, ensure that the programmes are embedded in an overall growth strategy for the country. 11 (d) Have in place clear procedures and, where appropriate, define KPIs to ensure that programme monitoring is systematic and accurately documented. (e) More comprehensively address data gaps from the outset. (f) Seek to reach an agreement with programme partners so that the roles and methods of co-operation are specified and transparent. (g) Better document the assumptions and modifications made to the economic calculations underlying the programme’s design. (h) Be more systematic in assessing the administrative capacity of the Member State to implement the reforms and the need for technical assistance. The conditions should be adapted to the results of this analysis. (i) Enhance its analytical work on programme design. It should, in particular, address the appropriateness and timing of measures, given the specific situation in the Member State. (j) Carry out interim evaluation for successive programmes of combined duration exceeding three years and use the results to assess their design and monitoring arrangements. (k) Analyse the appropriate support and surveillance framework for the period after the programme ends. 12 INTRODUCTION The three Greek Economic Adjustment Programmes 1. With the onset of the global financial crisis, the international environment changed drastically. The crisis required unprecedented bail-outs of financial institutions and other exceptional monetary and fiscal policies. The crisis was nonetheless followed by a global economic downturn and a debt crisis in Europe. Although countries with sound fundamentals managed to get back on a growth track within a relatively short space of time, countries with macroeconomic imbalances and structural weaknesses faced great difficulties. 1 2. In the context of the global financial crisis, the accumulation of macroeconomic imbalances, large stocks of public and external debt, weak external competitiveness, an unsustainable pension system and weak institutions all made Greece particularly vulnerable to increased risk aversion in the international markets. These factors, together with the disclosed misreporting of official statistics and the significant revision to the fiscal data for 2008 and 2009, negatively affected market confidence. As a result, leading credit rating agencies downgraded the sovereign ratings (from A-levels end of 2008 to C-levels in 2011). The market reacted strongly to this negative development, with the costs of financing for Greek debt increasing to unsustainable levels in the weeks before the country’s request for financial assistance. 3. As a result of losing market access at a reasonable price, Greece requested financial assistance from Euro area member states and the IMF 2 on 23 April 2010. Despite fiscal and structural measures undertaken by Greece, the first Economic Adjustment Programme (EAP) 1 As of 2008, besides Greece four other euro area countries (Portugal, Ireland, Spain and Cyprus) and three non-euro area countries (Latvia, Hungary and Romania) requested international financial support. 2 The International Monetary Fund (IMF) is an organisation of 189 countries and its primary purpose is to ensure the stability of the international monetary system. Among other actions, the IMF provides loans to member countries experiencing actual or potential balance of payments problems, which are conditional to the implementation of policies aimed at correcting the underlying problems. 13 was not sufficient for the country to return to market financing, and so two further programmes were agreed, in 2012 and 2015, respectively (see Table 1). Table 1 - The Greek Economic Adjustment Programmes Programme Facility Start End Total funding (billion euros) Euro area support disbursed (billion euros) IMF support disbursed (billion euros) Greece I GLF(1) 2010 2012 110 52.9 20.7 (17.541 SDR) Greece II EFSF 2012 2015 172.6 141.8 11.6 (10.2 SDR) Bridge financing EFSM 2015 2015 7.16 7.16 - Greece III ESM 2015 Ongoing (2018) 86 39.4(2) - 1 Bilateral loans coordinated by the Commission on behalf of participating Euro area Member States. 2 As of July 2017. Source: ECA based on Commission/ESM data. 4. The first adjustment programme was signed in May 2010 (see Figure 1 and Annex I). It was designed to run until June 2013 with funding of 110 billion euros (from the IMF and Euro area member states) intended to cover the economy’s financing needs and support the banking system (see Table 1). In exchange, Greece agreed to implement a structural reform programme covering economic, fiscal, financial and labour-market policies. The first programme ended prematurely in March 2012, as the unsustainable debt level and the deeper-than-expected contraction in economic activity revealed a need for further funding. By that time, 73.6 billion euros been disbursed. 5. The second financial assistance package was agreed with the Euro area member states and the IMF in March 2012 with total funding of 172.6 billion euros. The Euro area countries finally contributed 141.8 billion euros in the form of loans from the European Financial Stability Facility (EFSF) that had been created for this purpose. A key element of the second programme was the Private Sector Involvement (PSI), under which private investors contributed approximately 107 billion euros by writing down part of the value of their debt holdings. 14 6. Following the expiry of the second programme in December 2014 and its two extensions 3, and in view of the ongoing negotiations on the third package of financial assistance, Greece was granted a bridge loan from the European Financial Stabilisation Mechanism in July 2015. Given the uncertainty of the situation, in June 2015 restrictions in in the free movement of capital were introduced to safeguard financial stability (see Annex II). In August 2015, the Greek authorities and the Euro area partners agreed on the third programme proper, which took the form of a government loan from the European Stability Mechanism (ESM). The programme is scheduled to run until 2018. On 20 July 2017, the IMF Executive Board, at the request of Greece, approved in principle a precautionary Stand-by Arrangement of 1.6 billion euros, subject to further assurances. By September 2017 the IMF has not made any disbursements. 3 Respectively until end of February and end of June 2015. Figure 1 The chronology of the Greek crisis 15 4U The smut: .tn-n. oons'IderGreeroe in E.- 35 pawrent default AER-rent an End 31: U1 {eff I: Go'ternment is 3 PM resigns formed an hew coalition anternn-ent f/ isformed . F'I'i'tut'int'ions ututtl A LEIEMEI: on Athens Airport {ID-Elia and Daft Pimeus F'vorHiE I: 25 budget 2m: 'bu' retuIttu referendum Netu ?snl measure: I. Suspension of 5th - - II 2013. uete :nn-oun-oed Hit-1': cut on ?fulfil-L1: II budget ?g'eeTerIt :41 I if ISobernment 31:! :nn-ouoement . . 15 ?g'e-e'rent reshuffle hew :oalrtIon Go'ternment IE 1st 51F Let review Failure to form new :ustent'r goternment '5 go'ternen'ent matsurel: formed New pension l? Int-"Ina . system refvorm Ill euttujunt .4 2013' budget II-I 1U Statement on second buIqut ?Md I: 3915' bI-Ida? lay-toffsin 3 app-rented .5 litll'lE Pub?: Pi: chttuojunk chttuojunk Announoement that Lit. 5 new agreement is :l-mb in? ?st barn: needed Coalition barn: I: ursn'ent: 2r: bar-u: Announoement of gobernment Prrorichons 'sat'or referendum isforn'?l?i?fZ?lE Financial events ??at Economic?tdjustm err: Program ma Source: ECA. 16 Programme objectives 7. The main intervention logic of the Greek Adjustment Programmes has been to address the economic imbalances in Greece and thereby to prevent contagion between the Greek economic crisis and the rest of the euro area. To this end, the programmes aimed to establish a sound and sustainable economic and financial situation in Greece and to restore its capacity to finance itself fully on the financial markets. To achieve this objective, the Greek programmes4 focused on three key challenges which Greece committed to address through a comprehensive set of reform measures: (A) Restoring financial market confidence and fiscal sustainability: the programmes envisaged robust fiscal measures (including reductions in expenditure on pensions, the health system and public administration), as well as reforms of tax policy and tax administration. The fiscal reforms were intended to bolster confidence, regain market access and put the debt-to-GDP ratio on a sustainable path. (B) Safeguarding financial-sector stability: the programmes provided financial support to Greek banks via the Hellenic Financial Stability Fund (HFSF) aimed to address their urgent capital needs. The areas covered by the reforms also included the sector’s restructuring and consolidation, management of Non-Performing Loans (NPLs), governance, supervision and liquidity. (C) Promoting economic growth and restoring competitiveness: by means of structural reforms, the programmes aimed to improve cost-competitiveness and the overall business environment. This was intended to facilitate Greece’s transition towards a more investment- and export-led growth model. The areas covered by the reforms included labour and product markets, public administration, the legal system and education. 4 All three programmes shared the same objectives. Under the third programme, the modernisation of public administration was addressed under a dedicated objective, in recognition of the importance of the reforms in this respect, but the first and second programmes also addressed public administration reform under the objective “Growth and competitiveness”. 17 The programme partners and the role of the European institutions 8. The first two programmes were designed in the course of discussions between the Greek authorities and the so-called “Troika” consisting of the Commission, the ECB and the IMF (under the third programme referred to as “Institutions”). There were no formal arrangements defining the details of their cooperation. Under the third programme, in line with its treaty5, the ESM has also been involved in the design and monitoring of the programme. Although the IMF has not contributed to the financing of the third programme, it has remained fully active in the discussions on the scope of conditions and the adjustment path. 9. Throughout the programmes, the Commission’s role has been to act on behalf of the Euro area member states in close cooperation with other partners. Under the GLF, the Commission was responsible for coordinating and implementing the programme on behalf and under the instructions of the Euro area member states, providing support for, as well as negotiating and signing, a Loan Facility Agreement (LFA) and Memorandum of Understanding (MoU) on policy conditionality with Greece. The EFSF framework agreement and EU Regulation 6, applicable to the second programme, entrusted similar responsibilities to the Commission. The ESM Treaty called on the ESM’s Board of Governors to mandate the European Commission “to negotiate, in liaison with the ECB, the economic policy conditionality attached to each financial assistance”. 10. The ECB’s role throughout the programmes has been to act in liaison with the Commission to assess the policy conditions for financial assistance and to monitor on a regular basis the progress made on implementing those conditions. For the first programme, the European Council called upon the Commission as early as 11 February 2010 to monitor implementation of reform measures in liaison with the ECB. The EFSF framework agreement, 5 Treaty Establishing the European Stability Mechanism, paragraphs 12 and 13. 6 Regulation (EU) No 472/2013 of the European Parliament and of the Council of 21 May 2013 on the strengthening of economic and budgetary surveillance of Member States in the euro area experiencing or threatened with serious difficulties with respect to their financial stability (OJ L 140, 27.5.2013, p. 1). 18 the ESM Treaty and EU Regulation 7, defined the ECB’s role in a similar way for the second and third programmes. In this context, the ECB provided advice and expertise on different policy areas. Since the Single Supervisory Mechanism was set up (in November 2014), it has also participated in the discussions with the institutions and the Greek authorities. In addition, over the 2010-2016 period and irrespective of the programme process itself, the ECB published several opinions on draft legislative provisions submitted by the Greek authorities as these could have materially influenced financial stability8. The programming process 11. The financial assistance granted to Greece was subject to policy conditionality 9, anchored in the Greek Economic Adjustment Programmes and organised according to their main objectives (fiscal, financial and structural). The Greek authorities formally initiated the programming process by submitting the set of documents (see Box 1) to the lenders, updated for each of the reviews of the programme 10. The IMF and the European lenders remained independent when approving the programme documents and lending decisions (see Figure 2). Box 1 - The set of documents submitted by the Greek authorities under the first and second programmes - Letter of Intent – specifying key policy commitments, funding of the programme and further steps in the programming process. 7 Regulation (EU) No 472/2013. 8 Based on the Treaty on the functioning of the European Union (Articles 127 and 282) and Article 2 of Council Decision 98/415/EC of 29 June 1998 on the consultation of ECB by national authorities regarding draft legislative provisions (OJ L 189, 3.7.1998, p. 42). 9 Conditions to implement policy-specific reforms or legal changes. 10 For the first and second programmes the letters were addressed to the President of the Eurogroup, the Vice-President of the Commission and the President of the ECB. The same letter was sent in parallel to the IMF. 19 - Memorandum of Economic and Financial Policies (MEFP) – outlining the policies to be implemented by the Greek authorities, agreed primarily with the IMF and serving as a basis for compliance assessments by the IMF. - Memorandum of Understanding on Specific Economic Policy Conditionality (MoU) - specifying detailed economic policy measures, agreed primarily with the Commission that served as benchmarks for assessing policy performance as part of the quarterly reviews under the Euro area financial assistance programme. - Technical Memorandum of Understanding (TMoU) – setting out the definitions of indicators, assessment methods and information requirements. 20 Figure 2 - The programming process A Programme timeline Programme negotiation Decisions Compliance Compliance on granting report report financial assistance Decisionson Decisionson programme programme continuation continuation Council 7 I:l What are taken? decision Not oi! decisions are . Deficit targets needed for each review. Each Broad areas of structural For instance, if no decision reforms substantial changes are consists 01: three mode targets}, separate the Council decision does acts set in a not have to be amended. Memorandum of However, on Mot} update Understanding was systematically negotiated luv the approved after each European ?compieted? review. Commission and approved by Greece and the Eurogroup - Funding for the assistance programme Eurogroup approved by Eurogroup agreement:1 and Greece - Payments released subject to Commission and compliance reports Eurogroup agreement? For the first programme, the decisions were made as part of the Eurogroup unanimous agreements excluding Greece. For the second and third programmes, the decisions were tormallyir made by the EFSF and the ESM board of directors, based on the Eurogroup unanimous agreements excluding Greece. 21 C Sequence of steps Policy brief Negotiation or review visit EU law and Commission’s analysis Eurogroup position Defines the mandate for the Commission’s negotiation team Carried out in parallel with IMF and ECB staff EWG chair update to the + Interaction with other partners to align positions with a view to reaching a common agreement If changes to reform agenda are necessary Draft MoU EWG chair Commissioner Programme documents Explanatory note or compliance report + If Council conditions are to be changed Draft Council Decision Internal review by the Commission Decision by the College of Commissioners Agreement in the Economic and Financial Committee and Eurogroup Working Group Eurogroup agreement (depending on importance) Payment Signature of MoU (update) (if applicable) Source: ECA. negotiation mandate Council Decision Other partners IMF ECB ESM 22 12. The programme documents were drawn up as part of a negotiation process with the Commission, the International Monetary Fund 11 and the European Central Bank; this took the form of negotiation visits 12. For the Commission, the mandate for the negotiations was set out in advance in a policy brief which analysed the situation in Greece with respect to key policy fields and proposed measures for inclusion in the programme. For the purpose of the negotiations, Commission staff prepared the macroeconomic forecast underpinning the programme. Following the conclusion of each review, the Commission published it together with qualitative analyses of the macroeconomic and financial situation, the compliance assessment and the official programme documents. 13. Following the conclusion of the negotiations, the Council discussed the text submitted by the Greek authorities, based on the documents presented by the Commission 13, and issued a Decision under Articles 126 and 136 of the TFEU 14. After the Council Decision had been adopted, the Vice-President of the Commission (empowered by his fellow Commissioners) signed the final MoU on behalf of the lending member states (for the first programme) or the EFSF (for the second programme). The technical and legal details of the loan (such as the average maturity calculation, interest rates, disbursement and repayment 11 As of spring 2017, the IMF had not reached agreement on its financial participation in the third programme. It was, however, fully involved in the preparatory process. 12 The review missions were carried out by senior staff from the Commission, the IMF and the ECB. They aimed to: (1) update the forecasts and assessment of the country’s situation in key policy fields; (2) assess compliance with conditionality; (3) review policy conditionality for the next review. 13 Along with draft programme documents, the Commission provided the Council with a Communication by the Commission to the Council in compliance with Council Decision 2010/320/EU of 10 May 2010 establishing compliance with conditionality and Commission Recommendation for a Council Decision amending Council Decision 2010/320/EU (Articles 126 and 136). 14 As of the third review of the second programme (December 2012), no specific Council decisions were issued for the programme reviews as they did not lead to major changes in policy requirements. 23 arrangements) were specified in the Loan Agreement (for the first programme) 15 or the EFSF Financial Assistance Facility Agreement (for the second programme). 14. The third programme was agreed through a Memorandum of Understanding between the European Commission (which was acting on behalf of the ESM), the Hellenic Republic and the Bank of Greece. The third programme was subsequently approved by the Council. The financial terms and conditions for the loan were defined in a financial assistance facility agreement, and disbursement was approved by the ESM Board of Governors on the basis of compliance reports from the Commission (in liaison with the ECB). 15. The Eurogroup 16 provided political guidance for the process throughout the programming period and in particular reached agreement about the launch of the programmes, the main policy conditions, the financial allocation and the key financing conditions. The Eurogroup was supported by the Euro Working Group 17, which was consulted about the design of the programmes and decided to disburse the funds for the first and second programmes (the ESM Board of Governors for the third programme). AUDIT APPROACH 16. The audit examined whether the Commission’s management of the Greek adjustment programmes was appropriate. It complements our two earlier Special Reports, which assessed the Commission’s management of financial support to countries in difficulties under the Balance of Payment and European Financial Stability Mechanism (Special Report 15 The Loan Agreement was prepared by the European Commission, which managed the lending operations. For the first programme, the Loan Agreement was concluded on a bilateral basis between the member states of the Euro area and Greece. 16 The Eurogroup is an informal body where the finance ministers of the euro area member states discuss matters relating to their shared responsibilities related to the euro. 17 A preparatory body composed of representatives of the euro area member states of the Economic and Financial Committee, the Commission and the European Central Bank. 24 No 18/2015)18 and the delivery of Technical Assistance to Greece (Special Report No 19/2015). The audit addressed the following sub-questions: (a) Did the Commission have appropriate arrangements in place for managing the programmes? (b) Were the policy conditions appropriately designed and effectively implemented? (c) Did the adjustment programmes meet their main objectives? 17. The audit covered the Commission’s management of the three Greek Economic Adjustment programmes but, for the ongoing third programme, the audit focused on the design aspects. The design, monitoring and implementation of the programmes’ conditions were audited on the basis of a sample covering all key programme objectives. We also examined the Commission’s cooperation with the programme partners, but not the actions of the IMF 19 and the ESM 20. We also did not audit the actions of the Greek authorities, including the ownership issues which may have had an impact on the implementation process. We do not report on the ECB’s involvement due to the limitations of the audit scope (see paragraph 20). 18. We did not audit the EU’s high-level political decisions, for example on the justification to grant the financial assistance to Greece, and we limited the scope of the audit in several aspects. We did not consider the counterfactual scenario of no financial assistance. We also 18 This report presents a more detailed policy-specific analysis of the Greek Economic Adjustment Programmes compared to the audit work carried out for the five programme countries in the context of the Court’s SR 18/2015. However, the conclusions and recommendations of both reports are similar in nature, as DG ECFIN was managing the respective programmes in parallel. However we note that the Greek programme was exceptional in terms of scope and size and it presented a particular challenge for programme management. 19 The IMF’s involvement in the Greek Economic Adjustment Programme was assessed by the Independent Evaluation Office, “The IMF and the Crises in Greece, Ireland, and Portugal”, 2016. 20 The ESM is an intergovernmental organisation, funded by euro area member states. It is audited by a Board of Auditors (including one from the ECA), but as such the ECA has no rights to audit the ESM. ECA’s landscape review of EU accountability and public audit arrangements (2014) pointed out this problem. A recently published Evaluation report “The EFSF/ESM financial assistance” covered the institution’s involvement in Greece. 25 did not evaluate whether the Council chose the most appropriate financial allocations, deficit targets and fiscal paths to resolve the crisis. We did not assess the financing conditions of the loans granted to Greece, or the timeliness and appropriateness of the debt relief or the sustainability of the debt. We assessed the Commission’s macroeconomic projections and funding-gap calculations. When auditing the Commission’s cooperation with other partners, we did not assess whether such involvement was justified. 19. The audit evidence was collected on the basis of a detailed review of documentation relating to the EU’s financial assistance programmes (programme documents, internal Commission analyses, spreadsheet forecasts, and evaluations and studies by other organisations); a scorecard analysis of programme conditions and interviews with Commission staff; the national authorities, such as sector ministries, the Bank of Greece, business unions and stakeholders and the staff of the IMF, the ESM and the OECD. Scope limitation on the ECB’s involvement 20. In line with the ECA’s mandate to audit the operational efficiency of the ECB 21, we have attempted to examine the Bank’s involvement in the Greek Economic Adjustment Programmes22. The ECB questioned the Court’s mandate in this respect and the documentation we have been provided with by the ECB was insufficient evidence to carry out the audit work 23. Due to this scope limitation, we were unable to report on the ECB involvement in the Greek Economic Adjustment Programmes. 21 Article 287 of the Treaty on the Functioning of the European Union in conjunction with Article 27.2 of the Statute of the ESCB and of the ECB. The limitations of the Court’s audit mandate with respect to the ECB were addressed in our landscape review of EU accountability and public audit arrangements (2014). 22 The European Parliament specifically asked the Court to analyse the role of the ECB in financial assistance programmes (see European Parliament resolution of 27 April 2017 on the Court of Auditors’ special reports in the context of the 2015 Commission discharge (2016/2208)). 23 The ECB considers to have provided ECA with written answers and related documentation on its involvement in the Greek programmes, within the legal perimeter of ECA’s mandate to audit the operational efficiency of the management of the ECB. We encountered similar problems with obtaining evidence from the ECB in the audit of the Single Supervisory Mechanism (Special Report No 29/2016, paragraph 29). 26 OBSERVATIONS PART I – MANAGEMENT OF THE ECONOMIC ADJUSTMENT PROGRAMMES FOR GREECE 21. The management of the Greek Economic Adjustment Programmes involved a number of complex internal processes at the Commission which constituted a horizontal framework for policy-specific analysis and compliance assessments (see Part II). We can broadly distinguish three types of these horizontal processes: - Procedural arrangements for the design and monitoring of the programmes – defining the principles and mechanics of programme management as well as decision making procedures. - Co-operation with programme partners – division of responsibilities, communication and co-ordination of actions with the IMF, ECB and ESM. - Fundamental economic analyses – preparation of macroeconomic projections and funding gap calculations, setting key economic assumptions of the programmes. Arrangements for designing and monitoring the programmes’ conditions Urgency at the inception of the programmes 22. The design and institutional arrangements for the initial Greek Economic Adjustment Programme need to be viewed in the context of the urgency with which the Programme was set up. The immediate reason to apply for financial assistance was the need to borrow nine billion euros for a debt repayment due on 19 May 2010. Given that in spring 2010 market financing costs for the Greek debt increased to unsustainable levels, the Greek government needed to seek alternative financing in a short period of time 24 (see Figure 1 and Annex I for a detailed chronology of the crisis). 24 In January, the latest access to the markets provided 5 billion euros already at a high cost (6.2 %). Between January and April 2010, spreads for the 2-year bonds increased from 347 to 1552 basis points. 27 23. Given the anticipated difficulties in accessing market financing for short-term debt, on 25 March 2010 the Euro summit 25 agreed on the terms of the financial support to be provided to Greece. The Greek authorities submitted the official request for financing on 23 April 2010, i.e. less than one month before the disbursement had to be made. The first Greek Economic Adjustment Programme was signed on 3 May 2010. Given the depth and scope of imbalances in the Greek economy, this was a very narrow timeframe within which to design a comprehensive programme addressing all weaknesses. The design procedures and outcomes 24. The Commission’s general procedures for managing macro-assistance programmes were first established in April 2011, i.e. 11 months after the launch of the Greek programme and three completed reviews. The document defined internal arrangements for information and approval flows inside the Commission, with the Council and the programme partners. For example, it specified the steps for approving the main programme documents (policy briefs, compliance reports, draft MoUs and draft recommendations for Council decisions where relevant).The Commission also produced road maps for each financing facility, setting out the formal steps on the way to disbursement. 25. However, the procedures did not deal with the content of the programme’s conditions. There were no specific internal Commission guidelines on the design of the programme’s conditions in terms of scope, level of detail and the depth of the analytical work needed to justify them 26. 26. The key working documents underlying the programme conditions were the policy briefs. These policy-oriented documents were issued for each review and served as a basis for negotiating the programme reviews (with some degree of flexibility). The process for reviewing policy briefs was stipulated in the Commission’s procedures, but was not documented. 25 Euro summits are the meetings of the heads of state or government of the member states of the euro area. 26 The IMF has this type of guidelines (e.g. ‘Guidelines on conditionality’, September 2002). 28 27. On several occasions, Commission staff produced analytical documents supporting the policy briefs and corroborating the design of conditions in certain areas. However, this was not done systematically for all programmes and all policy areas (see Figure 3). Figure 3 - Documentation of the programmes’ design – examples Source: ECA. 28. We found certain problems in the design of the conditions, across several policy fields: • Several programme conditions were described in general, vague terms and were not measurable, in particular at the inception of the programmes. For example, in the field of export promotion, conditions such as “adoption of measures to facilitate public-private partnerships” or “measures to strengthen export promotion policy” did not specify any clear or concrete actions. In the financial sector, there was a condition in the second programme to establish an implementation plan aiming to improve collection of NPLs and establish targets in this respect in banks under liquidation by end-July 2013. The condition was open to interpretation and contributed to a significant delay in implementation. 29 • Conditions became progressively more precise. In the second programme their number and level of detail significantly increased. This was partially a result of implementation problems and delays. On the other hand, the high level of detail challenged the authorities’ administrative capacity and ownership, as conditions were not always sufficiently discussed and agreed at the design stage 27. • Increasing level of detail was combined with a lack of ex-ante clarity about the relative importance of each programme condition. Certain conditions were recognised as particularly important only ex-post when they were not implemented on time and then designated as prior actions during the programme reviews. Compliance with these conditions was explicitly required before a tranche of the loan was released 28. In the third programme, important conditions are now flagged as ‘key deliverables’. Overall growth strategy 29. The programmes had clear medium-term objectives, the primary expected result being the restoration of Greece’s access to market financing which necessitated the focus on fiscal consolidation. However, despite the Commission’s efforts, the programmes were not supported by an overall Greek-led growth strategy that could have extended beyond the programmes. A long-term strategy could have been more effective in coordinating the adjustment process and the design of measures in the various policy areas concerned. 30. The lack of a strategy was particularly pertinent in the field of export facilitation. The scope of the programmes’ measures in this respect – focusing on price competitiveness and the administrative burden – was not sufficient to favour a transition towards a more exportled growth model and did not address all the weaknesses identified by pre-programme 27 For example, there was a lack of consensus regarding the targeted level of autonomy for tax administration vis-à-vis the ministry of finance (in terms of work and resources planning and decision-making). Despite the increased number of conditions in this area, the programme objective was not met (i.e. the third programme deals with unsolved issues of autonomy). 28 The first programme contained no prior action in the area of tax administration, while 10 % of ‘unique’ conditions for tax administration in the second programme were prior actions. Three quarters of them were not initially included in the second programme. 30 analysis. In particular, the first programme did not propose a comprehensive strategy for sectors with a competitive advantage by taking into consideration the specific structure of the Greek economy. The sector-specific conditions were, however, included in the second programme. Cases of inconsistently designed measures and insufficient cross-policy coordination (see Box 2) show that the overall design of the Greek programmes also lacked an explicit strategy. Box 2 - Weaknesses of cross-policy coordination Product market reforms versus labour and tax reforms. Despite major labour market reforms and a sharp reduction in labour costs, prices did not fall sufficiently by 2013. This indicates that product market reforms failed to address market rigidities in the early stages of the programme. Furthermore, limited price adjustments partly reflected higher indirect taxes and charges associated with fiscal consolidation, thereby making structural reforms less effective in terms of their impact on prices and growth. The tax reforms in the adjustment programmes led to trade-offs and that in the first stages, the short term priority was fiscal consolidation, given Greece’s sizeable public deficit. Banks’ financial health versus fiscal measures. After banks proved to be underfinanced in 2012 and Greece’s sovereign rating had been fully discounted, there was no assessment of how fiscal measures would additionally affect banking debtors’ solvency and thus the market value of banks’ loans. For example, there was no analysis how higher recurrent taxes on property would hit the real estate prices and the housing loans. Business environment reforms versus tax reforms. There was no strategic roadmap to stimulate potential growth drivers in Greece, and this was reflected in the lack of a pro-growth fiscal consolidation strategy. There was no risk assessment of how potential/alternative fiscal measures (e.g. spending cuts versus tax hikes) and their sequencing in time would impact GDP growth, exports and unemployment. Political instability during the programme period 31. Another challenge that impacted the management of the programmes was a stop-start approach to the reforms, resulting to some extent from political instability in the country. From October 2009 until January 2015, Greece went through six elections and a change of government without elections in November 2011 (see Annex I). On each occasion, it took some time for the new leadership to reconfirm its commitment and approach to the 31 reforms, as well as to redefine the working arrangement with the institutions. The referendum held in July 2015 on the package of measures also contributed to the uncertainty of the programme’s future and to economic instability29. The lengthy political negotiations and uncertainty about the programme’s continuation hindered the smooth management of the programmes and implementation of the reforms. 32. The actual timing of the programme reviews largely departed from the initial quarterly schedule, reflecting implementation delays and difficulties in achieving an agreement on new measures (see Table 2). The most difficult situation in this respect was the transition between the second and the third programmes, with a gap of 16 months between the last concluded review and the approval of the new programme. While the second programme was never concluded 30, the transition involved a double extension of the second programme and a bridge loan from the EFSM. 29 Greece called a bank holiday because of the bank run and the queues which had formed at ATMs following the announcement of the referendum. 30 Despite protracted negotiations, no agreement was reached with the Greek authorities on the reforms needed to complete the last reviews of the second programme. 32 Table 2 - Reviews of the Greek Economic Adjustment Programmes Year Review Report date Review duration (months) 1st Programme May 2010 1st review Aug 2010 3 2nd review Dec 2010 4 3rd review Feb 2011 2 4th review July 2011 5 5th review Oct 2011 3 2nd programme March 2012 5 1st review Dec 2012 9 2nd review May 2013 5 3rd review July 2013 2 2014 4th review April 2014 9 2015 3rd programme Aug 2015 16 2016 Supplemental MoU June 2016 10 2010 2011 2012 2013 Source: ECA. Monitoring procedures 33. The process of monitoring compliance with programme conditions was reflected in the Commission’s general procedures for managing macro-assistance programmes, but only with regard to procedural aspects (i.e. the division of responsibilities and the process for approving the compliance report). However, there was no Commission procedure defining the key requirements on the scope, nature and documentation of monitoring. 34. The primary tool for monitoring Greece’s compliance with programme conditions, which was carried out jointly by the Institutions, was the “Compliance Table”. For the first and second programmes, the document was an integral part of the set of documents published after each review. The compliance table was completed and reviewed by Commission staff and subsequently presented to the Euro working group together with other programme documents. However, the review processes were not documented. In addition, the Commission officially notified the Council how it had monitored the conditions specifically referred to in the Council’s decisions. 33 35. The tables usually specified the compliance status for each programme condition requiring a specific review and brief justification (references to the legal basis, the state of play and reasons for any delays). However, not all conditions were systematically monitored in this way or on time. As the second programme has not been completed, there was no formal assessment of the conditions added in the last (fourth) review. The Commission followed-up developments in Greece, including through policy briefs, between the second and the third programme. The first compliance report for the third programme (published in June 2016) did not contain the compliance table, even though several conditions were due to be met by June 2016. Monitoring practices 36. For most of the conditions analysed, the Commission was able to demonstrate that it had assessed compliance with the programmes’ conditions and had a sound basis for doing so. In several policy fields, the Commission carried out in-depth reviews of the legal acts that were subject to the conditions and followed up on the necessary amendments. In some cases, the Commission used external expertise (e.g. in the field of regulated professions) to ensure a comprehensive assessment and duly amended the respective conditions in subsequent programme reviews. 37. Although the Commission regularly monitored and reported on formal compliance programme conditions, we occasionally found weaknesses such as missing or inaccurate assessments (see Annex III). Furthermore, the Commission’s compliance reports sometimes lacked clarity regarding the assessment of conditions. The fact that many different terms31 were used to report non-compliance with programme conditions introduced ambiguity as regards the overall meaning of the assessment. 38. In some cases, however, formal compliance with a condition (e.g. adoption of a specific piece of legislation) did not ensure effective implementation of that condition, i.e. it did not lead to the intended results. One example of this was the implementation of the Better 31 For example: ‘not observed’, ‘ongoing’, ‘not observed, progress made’, ‘observed and ongoing’, ‘partially observed’, ‘largely observed’ and ‘delayed’. 34 Regulation Agenda. After the respective law had been passed in 2012, the actual weaknesses still persisted, as the OECD noted in its Regulatory Policy Outlook 2015 32. 39. On the whole, although the Commission monitored the achievement of the programmes’ quantitative targets (i.e. a reduction in public employment, payroll costs and management posts), it had limited instruments in place to check that structural reforms were properly carried out. Difficulties with access to comprehensive data contributed to this problem. We identified particular data gaps in the area of taxation (see Annex IV, Part A) and regulated professions (incomplete data on the restrictions for key professions such as lawyers, notaries, engineers and architects). The programmes did not address the data gaps comprehensively, although there was some improvement in later reviews. For most of the structural conditions, the Commission was able to monitor the adoption of legal changes, but lacked adequate performance indicators and targets to assess the implementation of reforms. 40. The lack of adequate performance indicators hindered the Commission, for example, from verifying the achievement of efficiency objectives in the context of the reorganisation of the central administration. In the financial sector, only in April 2014 did the second programme introduce a condition that asked the Bank of Greece to set Key Performance Indicators to monitor banks’ progress in reducing their large NPL portfolios. Further, in the financial sector, the Commission did not fully use its capacity as an observer at the HFSF’s decision-making bodies, which provided an additional opportunity to ensure compliance with programme conditions 33. 32 For example, the Better Regulation Office was indeed formally established as part of the Agenda, but remained understaffed and had no power to decline draft proposals accompanied by poorly developed Regulatory Impact Assessments. 33 We found one particular case in 2013 where the Commission did not express convincingly its concerns in the HFSF regarding a potential merger between two banks. The transaction was overturned only later, but the delay resulted in a lost opportunity to recapitalise one of these banks partially with private money (it was eventually recapitalised in full by the HFSF). 35 Cooperation with programme partners 41. Conditionality at the technical level was designed mainly by the Commission (with coordination by DG ECFIN) and the IMF, in consultation with the ECB. Under the third programme, the ESM acted as a fourth institutional partner. The final MoU texts reflected agreement within the institutions, as well as between the institutions and the Greek authorities, which required all parties to adjust their positions where necessary. 42. There were no guidelines or specific procedures for cooperation between institutions and the process was not formally documented (e.g. in minutes) which impacted transparency of the process. However, the respective programme teams established informal cooperation arrangements. The teams were in regular contact by phone and email, as well as at meetings. The teams exchanged information, data and preliminary analysis, and discussed alternative options for programme design. The Commission and the IMF shared their draft programme documents to ensure conditionality was consistent. The fact that the staff of the three institutions’ had a variety of expertise and experience made it possible to produce more thorough assessments and forecasts, thereby reducing the risk of errors and omissions in programme design and updates. 43. The ECB and the Bank of Greece tailored some of their instruments to the programmes’ economic policies34. However, the programmes were not explicit about the extent to which the provision of liquidity to Greece was conditional upon fulfilment of the programmes’ conditions35 and how this support was taken into consideration for the purpose of estimating macro-economic forecasts and funding gaps. 34 For example, through the Securities Markets Programme put in place in May 2010 for sovereign bond purchases on the secondary market and the Emergency Liquidity Assistance put in place for solvent banks starting with 2010. 35 On 4 February 2015, the ECB decided to suspend the waiver for accepting Greek government bonds as loan collateral, thereby automatically increasing short-term borrowing costs for the banks. It was not clear whether the decision was made in coordination with the partners involved in the second programme. In the same month later on, the Eurogroup decided to extend the second programme by four months (until end of June 2015). 36 The programmes’ economic assumptions 44. Macroeconomic projections and the funding-gap calculation were the two key processes that defined the economic framework for designing and modifying the programmes (see Box 3). The Commission carried out these analyses for each programme and produced an update for each review. All institutions involved in programme management conducted their economic analyses separately, but the final result reflected the consensus they achieved. Box 3 - Key features of the economic processes underlying the programmes’ design Macroeconomic projections –provide the Commission’s best estimate of the main economic developments in the country, based on the assumption that the programme’s conditions will be fully implemented. The projections capture developments in the real economy (GDP and its components, and labour market developments), as well as the fiscal path (debt and deficit) as part of a Debt Sustainability Assessment. The projections entail a process that interacts with - but also goes beyond - the Commission’s usual forecasts. Even when the timing differs, forecasts under economic adjustment programmes benefit from the latest available European Economic Forecasts, which are used as input by staff. Funding gap calculation – estimation of Greece’s financing needs in order to ensure that the funding gap can be covered by the financial assistance committed under the programme. The funding gap is defined as the government’s financing needs, calculated as the difference between expenditure and revenue from sources other than programme assistance. The methodological strengths and weaknesses 45. The spreadsheets used for macroeconomic projections provided clear and detailed information on historical data and projections for the range of macroeconomic variables. Forecast trends for the main macroeconomic and fiscal variables were in line with the projections of their components and factored in a wide range of information sources. The Commission used the most recent assumptions on the dynamics of the international economy and technical indicators. However, the Commission did not explicitly incorporate some factors typically used in macroeconomic models and did not sufficiently integrate the macroeconomic and fiscal projections (see Box 4). 37 Box 4 - Weaknesses of the Commission’s forecasting methodology Comprehensiveness – certain factors (e.g. potential output and the non-accelerating inflation rate of unemployment) were not explicitly incorporated in the macroeconomic projections. The Commission also omitted to document how they were used in the judgment phase and which EEF elements were used for the programme forecast. Coordination of projections - The Commission carried out the macroeconomic and fiscal projections separately and did not integrate them into a single model. The mutual impact of both projections could therefore be integrated only on a judgemental basis. Approach to the Debt Sustainability Analysis - the long-term path of the debt expressed as a percentage of GDP was defined by a widely applied mechanical procedure of debt accumulation. This approach lacks an internal mechanism establishing an interaction between the debt path and economic activity (production, distribution and consumption of goods and services). 46. On the whole, the complexity of the funding gap methodology increased throughout the programmes and the spreadsheets for the second programme included much more complex calculations, valuations and modelling tools. The scope of the funding gap calculation was also extended to consider additional factors and elements later in the programme, though certain methodological weaknesses persisted (see Annex V). The methodology remained largely unchanged for the third programme. 47. At the Commission, there were no procedures for initiating, authorising, recording, processing and reporting the funding gap calculations. Quality controls were informal in nature and we identified significant weaknesses in the control environment for the spreadsheets 36. There were no improvements in this respect during the three programmes. 48. Despite the use of multiple sources, the data were not referenced or dated. This led to reconciliation problems and we could not verify whether the most up-to-date information had always been used. In the presentation of the first programme’s funding gap results we 36 For example, no change/version/access controls, no input reconciliations and in-built analytics to ensure coherence and data integrity. 38 identified a number of errors. They had no impact on the overall amount of the funding gap, but point out to insufficient Commission’s internal quality control arrangements37. Sensitivity tests and justification of assumptions 49. The Commission regularly monitored Greece’s liquidity situation and also updated the file used for the funding gap calculation between the reviews. The public programme documents did not present tests to establish how sensitive the funding gap calculation was to different outcomes or developments, but such analyses were carried out in the Commission’s working documents. The Commission built its scenarios around various elements, such as the schedule of different disbursements, the level of privatisation receipts and the arrears settled. On the other hand, no scenario analysis was carried out for the macroeconomic projections. The projections did not systematically incorporate external shocks to the Greek economy, such as a major depreciation of the euro or an abrupt decline in commodity prices, which actually materialised later during the programmes. 50. As regards justification of assumptions and documentation, we found weaknesses in both processes. For the funding gap, the programme documentation included a paragraph on the treasury management and financing of the programmes. However, the comments were often vague and did not explain the changes made to the funding gap calculation updates. In a number of cases, the specific assumptions underlying the funding gap calculation merited explicit justification 38. The justification problems remained in the third programme. The Commission agreed with a suggestion by the IMF that the targeted liquidity buffer should be increased to 8 billion euros. However, the Commission could not justify why this specific level for the liquidity buffer and referred to past experience from other programmes. 37 For example in the initial FG calculation the total debt roll-over needs were understated and in the December 2010 calculation the figures for the debt amortisation did not total. 38 For example, the estimation of public enterprises’ financing needs, reliance on accrual-based numbers rather than actual cash needs, conditions of market financing and the level of Stock Flow Adjustments. 39 51. When incorporating the impact of structural reforms in the macroeconomic projections, the Commission relied on the judgemental input, which was not documented. Further, because of the separated treatment of fiscal and macroeconomic projections, it was not possible to verify how the Commission defined and quantified the impact of fiscal policies on the macroeconomic projections. Accuracy and adjustments 52. Especially in the first programme, the Commission’s forecast significantly underestimated the depth of the crisis (see Annex VI). This was caused mainly by underestimating the drop in private consumption and investment. However, the overall precision of the Commission’s macroeconomic forecast was generally similar to those of other organisations39 and was partially influenced by the misreporting of statistical data for the pre-programme period by the Greek authorities. For 2010 and 2011 the Commission’s initial forecast underestimated the general government balance, as opposed to 2012-2013. 53. The funding gap calculations left Greece with very little room for manoeuvre as regards the liquidity buffer, defined in terms of cash position. On average, it decreased during the programming period, but without justification. The buffer was even projected to be zero at the end of the first quarter of 2014. The average cash position of 3 billion euros represented only 11 days of primary spending. Increasing the liquidity buffer would, however, necessitate adjusting the overall funding of the programme or introducing further fiscal measures in the programme. 54. Given the limited liquidity buffer and the deteriorating fiscal situation, Greece borrowed on a short-term by issuing T-Bills or delaying payments and increasing arrears. The amount of T-Bills to be issued varied significantly between the funding gap calculations and served as the main adjusting variable to keep the financing of the Greek economy on track. 39 We carried out in this respect a statistical study, comparing the Commission’s macroeconomic forecast for the first and second programme with the forecast of other organisations and alternative macroeconomic models. The comparative analysis, implemented through a forecast racing exercise, revealed overall that the precision of DG ECFIN forecasts perform was comparable to the forecasts of the competing institutions. 40 PART II - DESIGN AND IMPLEMENTATION OF THE REFORMS 55. Financial support disbursed under the three Economic Adjustment Programmes for Greece was conditional upon the implementation of reforms linked to the three programme objectives, i.e. fiscal, financial and growth-related. The specific design and scope of reforms has evolved throughout the programmes, but has been very broad from the outset. The programmes covered virtually all functions of the Greek state, in some cases requiring deep structural changes. There were substantial delays in compliance with programmes’ conditions across the policy fields (see Annex VII). 56. Our analysis, as presented in this section of the report, covered a sample of policies and focused on those with the greatest impact in terms of achieving the programme’s objectives. This part includes dedicated sections on taxation, public administration, financial sector and labour market reforms, while other policy-fields covered by our audit work contributed to the analysis in Parts I and III (see Figure 4). 40 Figure 4 - Audited policy fields 1 Results of the analysis considered in the cross-policy findings (parts I and III) Source: ECA. 40 Other key reform areas not covered by our analysis were the health system, education, the judiciary and industry. 41 57. For each of the policy areas, our audit work covered the following aspects: - the design of policy conditions (justification and analytical background of the measures, appropriateness, specificity and timing); and - implementation and results (compliance with policy conditions and the progress of reforms in the specific areas subject to the programme’s conditions). By contrast, the broader impacts of the reforms (in terms of growth, fiscal sustainability and financial stability) are discussed in Part III of the report to distinguish from the results of the specific conditions which are presented in the following sections of Part II. Taxation 58. Sustainable improvement of the Greek fiscal balance and boosting growth were among the main objectives of all three Economic Adjustment Programmes. These medium-term objectives were also shared by the IMF. To achieve them, the programme included conditions aiming to increase tax revenue and make tax administration more effective at collecting taxes (i.e. ‘reforms to enhance the government’s revenue-raising capacity’). In our audit work we focused on the changes in taxation system (including social contributions) and on reforms of the public administration to improve the collection of tax revenues. Conditionality design - tax policies Mix and stability of tax measures 59. General policy recommendations on fiscal consolidations (e.g. OECD) and pre-crisis assessments of the tax system in Greece considered that measures regarding taxes on consumption (indirect taxes) and broadening tax bases should be prioritised among tax measures due to their lower negative impact on economic growth. Contrary to these assessments, raising tax rates and temporary taxes had priority over broadening tax bases during the first programme due to their expected quick contribution to meeting the shortterm fiscal targets. 42 60. Measures to broaden tax bases were mostly part of the second and third programmes and some of them necessitated considerable reforms of the tax codes. The increases in tax revenue during the first two programmes mainly relied on indirect taxation as foreseen (in particular, indirect tax rates were increased and a recurrent broad-base property tax was introduced). However, certain indirect rate hikes (e.g. energy/pollution taxes) put additional pressure on the production costs of businesses. 61. The programmes’ conditions led to considerable instability in tax policy which affected all main types of taxes and not even the direction of the changes remained stable (for example the first programme advocated unified tax treatment of personal income sources, while the second programme moved in the opposite direction; see Annex IV, part B). To some extent, the instability was due to problems with administrative capacity in the area of tax policy design and weak implementation as acknowledged by the Commission, which necessitated frequent adjustments to the programme conditions. Under the first programme, the instability was also driven by the quest for ad hoc and urgent tax measures to meet deficit targets. The lack of an initial Commission’s strategy on tax policies and data gaps (see Annex IV, part A) also contributed to the problem, although improvement was visible throughout the duration of the programmes. Justification of the measures 62. The justification of tax policy conditions was not systematically documented 41. The important VAT and excise duties rate increases of 2010 were justified based on the need for fiscal consolidation and agreed with the authorities but there was no study of the consequences of alternative tax measures, such as a broad recurrent residential property tax 42, which was introduced in 2011 as part of an additional fiscal plan. The documentation at the design stage improved over time, based on better data from the relevant authorities 41 For example, the following condition was not justified: ‘Government changes legislation to mitigate tax obstacles to mergers and acquisitions such as the non-transfer of accumulated losses, together with the company and the complex computation of ‘excessive benefit’ (Law 3522/2006, Article 11) in the transfer of private limited companies.’ (first programme, May 2010). 42 The OECD considered this tax as the least harmful to growth. 43 (e.g. modelling of personal income tax reforms during the second programme and of VAT rate hikes of the third programme). Conditionality design - tax enforcement 63. Several international institutions (IMF, OECD, and Commission) flagged the considerable problems with tax compliance in Greece in the pre-crisis period. The first programme’s results in the area of tax enforcement were weak 43 and can be attributed to three broad types of problem: (i) the conditions were given limited priority, and were neither clearly specified, nor sufficiently sound (see Box 5); (ii) the implementation of the authorities’ own anti-evasion plans was weak; and (iii) the Commission’s monitoring was based on insufficient data and criteria (see Part I). Box 5 - Examples of problems in the design of tax enforcement measures Late inclusion in the programmes The conditions for the first programme gave limited priority to specific anti-evasion measures. A number of key measures were included only in the second programme (unification of taxpayers’ identification numbers, introduction of an IT system to interconnect all tax offices, establishment of a central register of bank accounts) or in the third programme (e.g. promotion and facilitation of electronic payments, publication of debtors for overdue tax and social security). Scope of the measures was not comprehensive Benchmarking with other member states shows that more specific measures could have been considered in Greece, e.g. a split-payment mechanism for VAT transactions with public institutions and more intensive use of electronic fiscal devices. Insufficient initial detail The initial first programme asked for ‘legislation to improve the efficiency of the tax administration and controls’, but there was no list of the specific agreed measures. The condition was based on IMF technical assistance recommendations of May 2010, but they were not fully transposed. 43 The Commission itself acknowledged weak progress with tax compliance during 2010 and 2011. 44 64. Most of the conditions for strengthening tax administration were included in the second and third programmes, thereby addressing some of the design weaknesses in the first programme. The conditions mostly relied on the recommendations of the Task Force for Greece (TFGR) and led to an improvement in terms of clarity, but this also significantly increased level of detail 44. This trend was explained by the fact that previous conditions had been implemented inadequately. Some conditions also had unrealistic deadlines (see Annex IV, Part C). Implementation and results 65. As a result of the tax-policy measures agreed in the programmes, the tax burden 45 on the economy increased during 2010-2014 (by 5.3 % of GDP) and continued to increase during the third programme due to additional tax measures. This shows that the implemented measures durably contributed to the improvement of the public finances. The hike was higher than in any other programme countries and made Greece reach the (nonweighted) average of tax burdens of the euro area member states at the end of 2014 (see Figure 5). Most of the increase was achieved during 2010-2012 which constituted the deepest phase of the Greek economic crisis. 44 There were 32 unique conditions in the MoUs for tax administration in the first programme and 194 conditions in the second programme. This large increase in level of detail is not reflected in the IMF conditions, which included 6 structural benchmarks in the first programme and 14 in the second programme in this area. 45 Taxes and social contributions revenue (as reported by Eurostat) excluding imputed social contributions as a share of GDP. By construction, the indicator excludes also uncollected amounts since Greece records taxes and social contributions with the method of ‘time adjusted cash amounts’. 45 Figure 5 - Tax burdens in Euro area programme countries (% of GDP) 41 38,5 39 37 35,4 35,4 35 36,6 33,5 32 33 31 35,9 30,6 Euro area average Greece 29 27 25 2009 2010 2011 2012 2013 2014 2015 2016 Source: ECA, based on Eurostat data (spring 2017). 66. By the end of the second programme, the Commission considered that Greece had implemented a broad range of conditions on tax evasion and collection, but their contribution to fiscal consolidation was limited 46. This was due to the fact that some of the reforms were not properly implemented, they still needed more time to be visible in statistics or their effectiveness was impacted by the instability of the taxation environment. Performance indicators of the programmes for tax administration show fragile and partial improvement in terms of the number of tax inspections and collection rates in the specialized audit centres, while the targets were rarely met between 2012 and 2014 (see Annex IV, Part D). During 2010-2014, tax debts increased by about 8.2 billion euros per year (4.2 % of GDP) on average despite write-offs and collection measures. The programmes did not set up any indicator for monitoring the estimated undeclared taxes (‘tax gap’) or undeclared work, but the estimated VAT gap 47 showed that evasion increased during the first programme and saw a fragile improvement in 2014 when compared with 2010. 46 In particular, VAT revenue as a share of GDP did not improve between 2009 and 2014 as initially forecasted despite several rate hikes and a sustained share of private consumption in GDP. 47 ‘Study and Reports on the VAT Gap in the EU-28 member states’ (DG TAX, 2016). 46 Public administration reform 67. The aim of Public Administration (PA) reform was to increase efficiency by cutting spending on public administration while at the same time increasing the quality of the public services offered. The programmes therefore comprised fiscal and structural components. The third programme placed further emphasis on the importance of public administration reform, which it established as a dedicated pillar of the programme. Our audit work focused on efficiency gains achieved through reform at the central level of public administration. Conditionality design The short-term priorities and strategic outlook of public administration reform 68. In an environment of fiscal crisis, the main priority of the first programme in the area public administration was fiscal consolidation 48. Structural conditionality gained momentum in the second programme, but was not based on adequate strategic planning. PA conditionality on structural reforms was not part of a broader, whole-of-government reform agenda approaching it sequentially in a coherent, planned way. Structural reforms of public administration did not address key factors which the Commission itself identified as necessary for the successful design of comprehensive public administration reform, i.e. an exchange of best practices at EU level, involvement of stakeholders and civil society, continuity and stability of reforms (see Annex VIII, Part A). 69. In 2013, the second programme asked the Greek authorities to present a strategy and an action plan for PA reforms, yet the requirement was not synchronised with key actions in the programme (the reorganisation of the Greek public sector started in 2012, preceding the delivery of the strategy by two years). This plan was not put into action and, in 2015, the third programme requested a new three-year strategy for PA reforms. 48 Wages and pensions in 2009 accounted for three quarters of all primary expenditure, having been the dynamically growing element of expenditure since 2000. The wage bill increased by almost 100 % in 2000-2008 and by a further 7.5 % in 2009. 47 The comprehensiveness of reforms 70. The first programme asked for a functional review of the central administration, to be performed by the OECD, to help prioritise public administration reforms and set up a customised, country-specific approach. While the review clearly identified the Greek administration’s weaknesses and those policy areas in need of administrative support, 13 of the OECD recommendations were included in the conditions after some delay (see Box 6). The recommendations were not systematically followed up, and there was no framework for tracking the progress of reforms or any strategy for addressing data collection and management issues (see Annex VIII, Part B). Box 6 - Examples of delayed implementation of OECD recommendations An ICT plan to secure interoperability between the various branches of the central administration was included in the programme one year after publication of the OECD report and, in the case of Human Resources and communication strategies, after two years. 71. Even before the first programme began, EU instruments were in place (primarily the Administrative Reform Operational Programme – AROP – financed from the European Social Fund) to address the weaknesses of Greece’s public administration. Although the Commission was aware of the depth of these problems, the programmes did not incorporate dedicated measures for developing institutional capacity 49. Some support in this respect, although of limited scope and scale, was coordinated by the TFGR50. By contrast, AROP capacity-building projects, which were already in place in 2010, were not linked with programme conditionality and some were cancelled during the programme period. The comprehensiveness of the conditions and coordination with other instruments (including technical assistance) improved under the third programme, in particular by establishing an explicit link between technical assistance and the programme. 49 Institutional capacity should typically cover three intervention dimensions for: (1) structures and processes, (2) human resources and (3) service delivery. 50 The size of and contribution by TA in the area of PA reforms was limited. It was financially supported by a 750 000 euro. However, this contract was under-executed (37 % of the available budget and 44 % of the total 620 man/days available for the 30.1.2013-1.1.2014 period). 48 Justification of the measures 72. Despite some analytical work by the Commission into the design of public administration reforms, the justification provided for certain specific measures was not satisfactory. In particular, the Commission could not provide any quantitative/qualitative analyses for the two key targets of the reforms (i.e. to reduce public employment by 150 000 employees in 2011-2015 and to complete mandatory exits of 15 000 employees by 2014 51). 73. The problem of insufficient justification also applied to the sequencing of reforms and deadlines. Although the Commission could not provide the analysis underlying the programmes’ design in this respect, the deadlines for certain conditions eventually proved over-ambitious and did not take implementation capacity into consideration. For example, the timeframe for setting up and implementing a new personnel appraisal system for the entire public administration was 11 months, while an OECD/SIGMA comparative study on PA systems in six EU countries suggested an optimal timeframe of 1.5 to 3 years for preparation of such reforms. Public administration staff were not subject to appraisal during the second programme. As a result, the third programme set a new condition, namely that a new modern performance assessment system should be legislated for within four months (by November 2015). The primary law was passed in February 2016, but additional secondary legislation needed to be introduced by June 2016 (seven months later). The first assessment will be performed by June 2017, almost two years after the start of the third programme. Implementation and results 74. The reforms of Greece’s public administration achieved the programmes’ objective to reduce costs. The target for reducing overall public employment was achieved in 2013, i.e. two years ahead of schedule, as a result of the attrition rule for new recruitment, reduced 51 The 150 000-reduction in the number employees was meant to be permanent and was implemented using the ‘one entry for five exits’ attrition rule, a decrease in temporary contracts, a labour reserve, pre-retirement and retirement schemes, and controls on hiring. For the 15 000 mandatory exits scheme, dismissed employees could be replaced by new staff whose profiles were better suited to the administration’s needs. 49 employment of contract staff and early retirement schemes. The public employment headcount dropped by over 225 000 between 2009 and 2015 (see Figure 6). Figure 6 - Employees in the core public sector and general government wage bill, 2009-2015 Source: European Commission, Apografi Database, Eurostat. 75. In terms of fiscal results, the annual cost of public employment fell by 37.9 % (9.6 billion euros) between 2009 and 2015. By 2015, the level of public-sector wages was aligned with the euro area average (9.1 % of GDP in the euro area v. 9 % in Greece; see Figure 7). However, these savings resulted in increased expenditure on the pension system. Figure 7 – Public employee compensation as a percentage of GDP 11.5% 11.0% 10.9 10.5% 10.3 10.0 10.0% 9.7 10.1 9.5% 9.9 9.4 9.6 9.5 9.0% 9.3 8.5% 8.0% EU 28 Greece Source: Eurostat data, payable net of social contributions. 9.3 9.2 9.1 9.0 50 76. The structural components of public administration reform faced limited support at the national level and were adversely impacted by the political instability (see Paragraphs 3132). Consequently, they brought a lot fewer tangible results than fiscal measures. The first programmes made limited progress in terms of reorganising public entities, Human Resources Management, and the de-politicisation of the civil service (see Box 7). By contrast, positive developments included the creation of a Governmental Council for Reform, the General Secretariat for Coordination of Government Work, the Single Payment Agency and a census database for public employment. Box 7 - Public administration reform – structural aspects Evaluation of public entities - the July 2011 review of the first programme required that, following a first wave of 153 entities, further evaluations should cover 1 500 public entities, including a comprehensive benchmarking procedure, although no action has been taken in this respect. HR management and de-politicisation - measures (selection of managers and a personnel appraisal system) were introduced late (2013) in the second programme and were not implemented. Depoliticisation of management and top management, including Secretaries General and personnel of entities of public and private law, was introduced under the third programme. Wage grids - the objective under the first programme was to reduce the fiscal burden, while only the second programme introduced conditionality to better reflect skills and responsibilities of staff, it was, however, not implemented. In 2015, the third programme asked for a new reform of the unified wage grid. The reform intended to modify the wage grid - in fiscally neutral manner - to better reflect the skills and responsibilities of the different jobs. Financial sector reforms 77. Safeguarding domestic financial stability was one of the main pillars of all Greek Economic Adjustment Programmes. To achieve this objective, the programmes included conditions to strengthen the health and resilience of the financial sector by enhancing bank capital, maintaining sufficient liquidity, tackling non-performing loans (NPLs) and improving governance and supervision (see Box 8 and details on the programme strategies and key conditions for the financial sector in Annex IX, Part A). Our analysis focused on the measures 51 taken to restructure and recapitalise the banking sector and the reforms relating to governance and asset quality, in particular NPLs. Box 8 - Key aspects of the financial sector reforms Hellenic Financial Stability Fund (HFSF) - establishment of the HFSF with the aim of contributing to the financial stability of the Greek banking system by providing equity capital, if necessary. Bank recapitalisation and resolution – an amount of 10 billion euros was earmarked under the first programme for bank capital interventions via the HFSF, but in the end it was only marginally used. Under the second programme, an amount of 50 billion euros was earmarked for bank recapitalisation and resolution costs via the HFSF. An additional buffer of up to 25 billion euros was envisaged for the same purpose under the third programme. Liquidity - preservation of sufficient liquidity in the banking system in compliance with Eurosystem rules. In the medium term, achievement of a sustainable bank funding model by gradually reducing banks’ reliance on central bank borrowing. Non-performing loans (NPLs) - strengthening the insolvency framework and banks’ NPL management capacity, development of a market for NPL sales and servicing, and enhancement of the management of banks under liquidation. Governance and supervision - strengthening governance of the HFSF, banks and the Bank of Greece. Enhancements in domestic supervision and regulation. 78. Notwithstanding the weak loan demand over the programme period, all programmes provided a very limited number of financial-sector reforms52 to directly address the issue of credit contraction in Greece either in terms of bank-credit supply or credit cost. Constraints on credit availability have hampered efforts to restore the competitiveness of the Greek economy, as domestic SMEs have faced significant difficulties gaining access to credit. 52 Although the programmes laid down conditions for improving SMEs’ access to finance, financialsector reforms only provided for the establishment of the Institute for Growth. In addition, the amount granted for this purpose (around 200 million euros) was small compared to Greek businesses’ financing needs. 52 Conditionality design 79. The Commission’s analysis did not always sufficiently justify the risks of some policy options. For example, in the context of the 2013 bank recapitalisation, the Commission considered several options for attracting private investors. As it concluded that the Greek authorities’ proposals offered excessive sweeteners to private investors, it finally opted for a warrants issue 53. However, the analysis did not comprehensively identify some of the risks of the warrants, e.g. their potential impact on bank share prices (see Annex IX, Part B). Restructuring and recapitalising the banking sector (i) Capital needs assessments 80. The capital needs assessments that had been envisaged in the first two programmes not always used sufficiently prudent assumptions. For example, even the adverse scenario for GDP growth, dated January 2011, was more positive than the actual economic performance by 4.6 p.p., 7.8 p.p. and 4.7 p.p. for respectively 2011, 2012 and 2013 54. In addition, the remainder of the second programme funds (around 11 billion euros) partially influenced the Commission’s analysis of a follow-up capital needs assessment conducted in early 2014 by BoG. (ii) Viability assessment 81. A condition of the second programme asked the Bank of Greece (BoG) to assess which of the 15 domestic banks were viable and thus eligible for recapitalisation using programme funds. The remaining banks would be resolved, unless private capital could be raised. The programme did not require the BoG to ensure that the process was sufficiently transparent. 53 A warrant is a security that entitles the holder to buy the underlying stock of the issuing company at a fixed price until its expiry date. 54 Since the onset of the crisis Greek banks have undergone six capital needs assessments. These are usually conducted on the basis of two scenarios (baseline and adverse) about changes in key macroeconomic variables. It should be noted that, in all cases, the Commission had an essential role in the scenario design (see also ECA Special Report No 5/2014, paragraph 58 and Box 1).The IMF has identified weaknesses in the capital needs assessments in its ex-post evaluation for the second programme (Country Report No 17/44, February 2017, paragraph 39 and Figure 19). 53 For example, the banks had no access to their own assessment and its underlying calculations. (iii) Bank recapitalisations of 2014 and 2015 82. The conditions in the second and third programmes allowed the HFSF to be used only as a last-resort source of support in the recapitalisations of 2014 and 2015 (see Annex IX, Part B). This meant that the HFSF could not participate in the bank recapitalisations in the event of private-sector interest in order to minimise further injections of public funds. In the recapitalisation process, significant downward pressure was thus exerted on share prices, which resulted in the substantial dilution of the HFSF 55, which was already a majority stakeholder as a consequence of the 2013 recapitalisation. Asset quality – Non-Performing Loans 83. In the case of NPL management, the creation of an asset management company (AMC) is one of the most widespread solutions at the international level. However, such a solution was not used in the Greek programmes due to funding constraints and other factors, such as the diversification of non-performing loans across all sectors, governance issues and EU state-aid considerations. Notwithstanding the potential constraints, the Commission’s estimates of an AMC’s financing needs were not concrete enough to provide a comprehensive assessment of whether the measure was appropriate in the Greek context. 84. The first two programmes placed emphasis on recapitalising banks and providing liquidity support, but this was not accompanied by effective action on NPL management (e.g. the first programme only provided a commitment to review insolvency legislation). The second programme attempted to address the problem of NPLs by relying mostly on banks’ internal NPL management which, although essential, proved largely ineffective 56. As the 55 In 2014, private investors obtained shares in the four largest banks at a discount of between 7 % and 23 % over the last closing price. In 2015, the discount was considerably higher, i.e. between 34 % and 93 %, as no minimum price was determined ex ante due to the high level of uncertainty. 56 The problems included implementation of the Code of Conduct on the management of NPLs being postponed until late-2015, less active loan portfolio management (banks focused mostly 54 second programme progressed, banks’ NPL management capacity was identified as a higher priority, but tangible reforms to improve internal management processes were put in place only as from the third programme. 85. With respect to business and personal insolvency in Greece, the conditions in the first and second programmes remained general in nature and the series of legislative amendments that were duly introduced to deal with distressed assets were mostly of limited-scope and ineffective 57. Many reforms were characterised by a piecemeal approach and a diverse set of policy priorities that often led to conflicting policy objectives. In some cases, they were even taken without proper stakeholder consultation and impact analysis. The complexity of the insolvency framework has perpetuated the NPL problem rather than addressing it within a coherent, centralised strategic framework with clear policy priorities. The framework’s design also did not take account of the judicial system’s limited capacity for dealing effectively with the large volume of cases. Taken together, such complexity and capacity problems have generated backlogs, thus disseminating strategic default and moral hazard throughout the system (see Box 9). on short-term solutions, thereby shifting the problem into the future), and poor interbank collaboration. Inaction was also observed in cases where banks had denounced loan agreements for which legal action had not been initiated. 57 Both the ECB and the IMF have identified several obstacles to NPL resolution (ECB, ‘Stocktake of national supervisory practices and legal frameworks related to NPLs’, Annex III, September 2016 and IMF, Country Report No 17/41, February 2017, p. 3-15). 55 Box 9 - Personal insolvency The personal insolvency legal framework resulted in around 200 000 applications flooding the judicial system, with debtors being granted a payment moratorium because of a huge backlog that is expected to last up to 15 years. The problem was exacerbated by insufficient guidance for judges and design problems that preceded the 2015 law (a programme condition of the third programme), such as the requirement that debtors pay a minimum amount or nothing at all pending a court hearing. Consequently, the framework was not well targeted, as ineligible debtors applied for the insolvency scheme anyway as a way of avoiding creditor actions until such time as their cases were heard. Nontargeted household insolvency rules also contributed to undermining payment culture among borrowers. The Bank of Greece and domestic banks estimate that one sixth of businesses and at least one quarter of households are strategic defaulters. 86. Only the third programme included a condition to create a functioning market for NPL servicing and sales. However, several key impediments were not removed either by the law adopted in late 2015 or by two further amendments in 2016 (also covering performing loans), and the legal framework remained burdensome. 87. In the case of banks under special liquidation which handled mostly non-performing loans, operations were fragmented for a long period, giving rise to low efficiency in terms of NPL collection and operating costs. For example, the consolidation of operations (16 entities into one) was only a condition under the third programme and was implemented more than three years after the original proposal by external consultants. The liquidation procedure absorbed 13.5 billion euros from the programmes’ funding via the HFSF (see Table 3) and further 1.7 billion euros from the national deposit guarantee fund. Governance (i) Bank governance 88. While financial-sector reforms gained in importance over the course of the first programme, the programme itself did not include conditions to enhance bank governance. However, governance-related problems (e.g. lending to related parties on non-market 56 terms) existed well before the crisis, as Greek banks’ corporate governance was, on average, considerably inferior to that of their European counterparts from the outset. 89. The second programme provided that the four largest Greek banks (‘systemic’) would be recapitalised mostly by programme funds via HFSF but without ensuring sufficient scrutiny of their private management. Contrary to international practice, changes of ownership that led in 2013 to the almost full nationalisation of the domestic banking sector were not reflected on most bank boards. In fact, in some cases, management control remained with historical shareholders and the HFSF was not entitled to assess them in terms of their experience, reputation and independence. A condition on the evaluation of banks’ corporate governance was only included in the third programme when the HFSF had to review all bank board members. 90. To this end, and in order also to tackle potential political or business interference, the third programme provided for stricter selection criteria as regards the qualifications and experience of bank board members. However, the criteria restricted the candidates to banking and financial expertise; such a requirement was not fully aligned with international practices and EU/SSM requirements, which, in principle, promote board diversity and collective knowledge. 91. Financial-sector reforms did not focus sufficiently on the governance and domestic supervision of less significant banks, either. Almost six years after the first programme was introduced, an on-site inspection by the BoG and the SSM in March 2016 revealed severe internal deficiencies in terms of governance, risk management and lending practices at one bank. (ii) HFSF governance 92. The first programme designed the HFSF’s original governance structure, which did not lead to enhanced independence from the authorities. Despite the second programme’s condition that the HFSF should have a two-tier management structure (i.e. consisting of the General Council and the Executive Board), independence issues persisted. The third programme addressed this weakness by focusing on the selection process for the HFSF’s senior management. However, the solutions proposed by the second and third programmes 57 did not ensure an efficient division of responsibilities and powers between the two decisionmaking boards. The HFSF’s decision-making process also raised serious concerns about its transparency. For example, in 2013, the HFSF approved the sale of a majority stake in a bank subsidiary even though the transaction was not based on a competitive tendering process with multiple bidders. 93. The programme conditions led to frequent changes in the HFSF’s senior management (34 people in the first six years, including four chairmen and four chief executive officers), a practice that entailed a risk of knowledge gaps and diminished influence in the banks in which HFSF held shares. Expected financial results 94. As of December 2016, the HFSF had injected 45.4 billion euros into the Greek banking sector, with expected losses of 36.4 billion euros (see Table 3). Overall, only a small part of the expected losses could potentially be recovered over time due to an appreciation in the market value of shares in systemic banks, and most programme funds for domestic banks are expected to remain part of Greece’s public debt. Table 3 – Expected results of intervention in the Greek banking sector (in billion euros, 2016) Funds used Funds recovered Estimated recoverabl e amount Expected losses Systemic banks 31.91 2.72 3.83 25.4 Banks under liquidation 13.5 0.5 1.9 11.0 Total 45.4 3.2 5.7 36.4 1 Including the funds used for capital injection of two temporary credit institutions (bridge banks). Including the proceeds from warrants, the redemption of CoCos and the respective interest. 3 Including the fair value of bank shares and the fair value of CoCos. 2 Source: HFSF draft financial accounts, December 2016. 58 Labour market reforms 95. The overall aim of the labour market reforms anchored in the programmes was to improve the performance of the labour market, among others by fostering a rapid adjustment of labour costs. The labour market reforms were expected to contribute to restoring cost-competitiveness in the Greek economy and containing job losses during the recession. With respect to labour costs, the second programme established an indicative target of a 15 % decrease in unit labour costs (ULCs) in 2012-2014. 96. Since their inception in 2010, the programmes have included a broad range of measures to address the rigidities of the Greek labour market. They focus on bargaining and wage formation, minimum wage setting and level, employment protection legislation, and working time flexibility. In parallel to liberalisation measures, the emphasis from the second programme onwards has increasingly focused on the fight against undeclared work and on active labour market policies aimed directly at job creation, but these aspects were not covered by the audit. Conditionality design The justification and analytical underpinning of the reforms 97. Policy briefs and other working documents substantiated the Commission’s approach to labour market reforms and provided analysis of alternative options (e.g. how to reduce unit labour costs or shape the wage bargaining framework). However, the Commission did not cover some risks in these analyses, especially in the early stages of design. In particular, there was no assessment of how the reforms could impact the grey zone of the labour market, although some of the measures entailed an increased risk in this respect 58. The social impact was assessed only for the third programme, but the analysis was descriptive in nature and there was no attempt to estimate impacts or burden sharing quantitatively, either for specific measures or for the entire set of labour market reforms. 58 E.g. reducing the minimum wage increases the incentive to pay part of the salary (exceeding the minimum wage) informally in order to reduce taxes and social security contributions. On the other hand, there were dedicated measures in place to address the informality of the labour market. 59 98. The Commission documented extensive exchanges of information and policy discussions with stakeholders (including the Greek authorities, business associations and other non-governmental partners), the IMF and internally. As regards the compromise reached by the Troika partners, the Commission’s view prevailed in some cases (there was no imposed wage cut in the private sector as advocated by the IMF), while the IMF prevailed in others (e.g. employment protection legislation was liberalised further after the first wave of reforms). In the latter case, the Commission could not fully justify deviation from its initial position. Appropriateness of measures 99. In some cases, the design of specific conditions also failed to take account of specific features of the Greek economy, in particular the prevalence of micro and small enterprises. For example, the reform of wage bargaining in its initial design provided an opportunity for bargaining at company level, although it still required formal staff representation. As small enterprises do not often have such an arrangement, they could not take advantage of the reforms. The issue was addressed only later in the first programme with the introduction of “unions of persons” 59, which allowed wage negotiations to take place directly with employees at company level. Timing 100. Labour market reforms were initiated in 2010 and the Greek authorities implemented the first wave of reforms (wage bargaining and employment protection legislation) shortly after the programme was approved. However, although several programme measures were implemented in 2010-2016, the same themes were present in the negotiations for the third programme. These developments point to problems with the timing and sequencing of reforms, which go beyond delays in implementation (see Box 10). 59 A form of employees’ representation in the absence of unions. 60 Box 10 - Timing of labour market reforms Some elements that were key to the effectiveness of the wage bargaining framework were included only in the second programme (e.g. unions of persons were established and automatic extensions of expiring agreements were limited if no compromise on renewal could be reached). The deadline for a first nominal reduction in the minimum wage was 2012. The new framework for setting the minimum wage was to be applied only after the second programme expired. The deadlines for implementing reforms were unrealistically short in some cases (mainly under the third review of the second programme), e.g. the review of labour regulations and compilation of all existing labour legislation have the same deadline. Implementation and results 101. In numerous cases, the implementation of labour market reforms was delayed compared to the programmes’ deadlines (by up to 21 months) or did not fulfil them at all (see Annex VII). These conditions were linked to key aspects of labour market reforms (e.g. wage bargaining and EPL). 102. Despite implementation weaknesses, we noted some improvement in indicators pointing to increased flexibility and competitiveness of labour markets: - Reduced restrictiveness of employment protection legislation. However, in 2013 (i.e. after key reforms were introduced) the indicator was still above the OECD average (see Figure 8). - A lower minimum wage – decreased from 863 euros in 2010 to 684 euros in 2012. - Significant adjustment of unit labour costs. The 14.1 % decrease in ULCs in 20102015 was the highest among EU member states. However, the second programme aimed to reduce the ULC in 2012-2014 by approximately 15 %. This indicative target was missed because during the respective period the ULC adjusted by 10.9 %. 61 Figure 8 – Compound indicator of the restrictiveness of the Employment Protection Legislation 13,00 12,50 12.07 12.07 12,00 11,50 11.11 11,00 10.78 10.70 OECD 10,50 10,00 9,50 Greece 9.47 9.50 9.46 9.39 9.31 2009 2010 2011 2012 2013 9,00 Source: ECA, based on OECD data. PART III - ACHIEVEMENT OF PROGRAMME OBJECTIVES 103. Seven years, three programmes and over 265 billion euros of disbursed assistance (more than 150 % of Greek GDP) reflect unprecedented dimensions of joint Euro area and IMF support for Greece. These figures alone show that the programmes were not successful in correcting deep economic imbalances or in enabling Greece to manage its financial obligations without external support. However, they provided for a continuity of funding and economic activity in Greece. 104. The Greek Economic Adjustment Programmes had objectives in three areas: - fiscal sustainability; - financial stability; and - return to growth. 105. In the course of seven years, the Commission has not attempted to evaluate in a comprehensive manner the extent to which either the first or the second programme achieved their objectives through mid-term or ex-post evaluations. Carrying out such evaluations was usual practice for other support programmes managed by the Commission and, in providing useful lessons learned, would be particularly relevant in the case of Greece, where three sequential programmes were implemented. 62 106. This section of the report analyses the achievement of the three main objectives of the programmes, wherever possible by applying the specific targets anchored in the programmes themselves. In this analysis we go beyond the assessment of the results of the specific measures (presented in Part II) and focus on the broader impacts which cannot be linked to single conditions of the programmes. Such analysis is helpful to understand the broad economic developments in Greece under the three Economic Adjustment Programmes, although it should be stressed that the developments in question were influenced not only by the programmes’ design and implementation, but also by a number of other factors. These included political uncertainty, which led to delays in implementation of reforms, impacted market confidence, and consequently reduced the effectiveness of the programmes. Ensuring fiscal sustainability 107. The Greek economic crisis started as a debt crisis, and the restoration of fiscal sustainability – including the country’s ability to finance its needs on the market – was the immediate objective of the programmes. Access to markets 108. The first programme envisaged that Greece would regain access to the financial markets in 2014 and finance its debt obligations for a total of 60 billion euros in 2015. In 2014, the Greek state financed itself on the financial markets on two occasions, although only for a total of 4.5 billion euros. This amount accounted for less than 5 % of the total financing of the second programme (110 billion euros). Two further assistance programmes were needed (with the third expiring in 2018) because the country had not regained the ability to finance all of its needs on the market. Use of funds and post-programme financing 109. Based on the most recent update of the funding gap carried out by the Commission for the first review of the third programme, Greece is expected to achieve a primary surplus of 1.4 billion euros in 2017 and 3 billion euros between January and August 2018, the last 63 months of the third programme. Therefore, the programme funding for this period will be used to repay the debt and will allow an increase of the liquidity buffer (see Figure 9). Figure 9 – Greece`s capital needs (billion euros) Source: ECA. 110. In the immediate post-programme period Greece will have to repay significant amounts of debt (see Figure 10). In 2019 the gross financing needs amount to 21 billion euros in capital payments and interest. The programmes’ assumption is that the after its expiry the debt repayment will be fully funded by the primary surplus and market financing. For the 64 second programme, some analytical work in the possible post-programme framework was done in 2014, while for the third programme this work is only commencing and therefore is not reflected in the programme documents. Figure 10 – Greece’s debt due (capital payments; billion euros) Source: Bloomberg. Fiscal consolidation 111. The fiscal situation at the onset of the crisis required a very deep adjustment. In 2009, the general government deficit exceeded 15.1 % of GDP, as compared to the 3 % threshold under the Stability and Growth Pact. The programmes’ main annual fiscal targets (i.e. the nominal deficit 60) were set by Council Decisions as part of Greece’s excessive deficit procedure (EDP). The programmes set further detailed nominal fiscal performance criteria in the MEFP, yet the fiscal plans that supported these targets were not always credible (see Box 11). 60 In the first programme, the Council Decisions referred to the headline deficit and in the second programme to the primary deficit excluding banking measures. 65 Box 11 - Credibility issues in the initial fiscal plan The initial fiscal programme (May 2010) aimed to achieve an unprecedented primary surplus of 5 % of GDP in 2014 (i.e. beyond the programme’s initial three–year duration). The fiscal programme estimated unspecified measures totalling 4.4 % of GDP (34 % of the overall fiscal effort) which were concentrated in 2013 and 2014. The Commission’s quantification of the VAT base broadening measures differed from the IMF’s estimates: - the IMF forecasted a fiscal yield of 1.8 % of GDP in 2013 stemming from structural fiscal measures (such as improved tax administration and better budget control and processes). This yield corresponded to unspecified measures in the Commission’s estimates of the 2013 measures; - unlike the Commission, the IMF did not estimate any fiscal measures for 2014. 112. Fiscal targets for the first and second programmes were only met in 2012 and 2013 (see Table 4). For the other years, they were missed by between 3 to 6 billion euros, despite the fact that Greece resorted to one-off revenue-increasing measures in order to meet either deficit or cash projections. During the first programme, the unchanged EDP nominal targets were not met mainly due to the worsening economic environment (i.e. unexpected drops in nominal tax revenue and hikes in interest expenditure). Tax base erosion partly wiped out revenue gains from important tax measures: this was typically the case for VAT revenue. Moreover, compliance with fiscal targets was supported throughout the programmes by the accumulation of government expenditure and tax-refund arrears. 66 Table 4 - Compliance with updated EDP targets 1st programme Council Decision (May 2010): headline balance (billion euro) Council Decision (Mar 2012): primary balance, excl. banking measures (billion euro) Council Decision (Dec. 2012): primary balance, excl. banking measures (billion euro) 1 2nd programme 1 1 3rd programme 2010 2011 2012 2013 2014 -18.5 -17.1 -14.9 -11.4 -6.4 -2 3.7 9.4 -2.9 0 2.8 Council Decision (Aug. 2015): primary balance (% GDP) Actual (first ex post monitoring) Actual (latest ESA95 or ESA10) Difference (applicable target vs. actual) 1 2015 2016 5.7 9 -0.25 0.5 -22 -20 -2.6 1.5 N.A. -3.4 3.9 -24.5 -20.1 -2.6 0.6 N.A. -2.3 3.9 -6 -3 0.3 0.6 N.A. -2.05 3.4 Applicable target in bold. Source: ECA. 113. Although Greece did not comply with the nominal fiscal targets, the primary structural balance 61 improved significantly. The overall adjustment exceeded 17 % of GDP in 20092015. This reflected Greece’s very deep imbalance in this area at the onset of the crisis. 114. In the context of the adverse macroeconomic situation, overall public debt continued to grow throughout the programme period, despite the fiscal effort made. In 2016, public debt in Greece exceeded 181 % of GDP compared to 126.7 % in 2009 (see Figure 11). The only year when the public debt fell was 2012, owing to the PSI procedure under which private investors were asked to agree to write off 53.5 % of the face value of their Greek government bonds. The first programme projections estimated that public debt would peak in 2013 (at 149.7 % of GDP under the initial first programme and 164.2 % of GDP under the 61 Primary structural budget balance: The actual budget balance net of the interest expenditure, the cyclical component and one-off and other temporary measures. The structural balance gives a measure of the underlying trend in the budget balance. 67 second. The strongest contributing factor to the relative increase of the debt was the nominal GDP decline. Figure 11 – Public debt to GDP ratio Source: European Commission. Sustainability of the pension system 115. One of the key impediments to stabilising the fiscal situation in Greece remains the considerable cost of the pension system. Successive cuts in pensions reduced the nominal expenditure on pensions by 2.5 billion euros in 2009-2015. However, due to the GDP decline, the pension costs as a percentage of GDP rose systematically in 2010-2015 thus showing the limited effectiveness of the reforms in terms of sustainable fiscal adjustment. In 2016, public expenditure on pensions was the highest in the euro area (over 16 % of Greece’s GDP), while annual state funding of pension-system deficits exceeded 9 % of GDP (the euro-area average was 2.5 %). 116. Based on the projections of the 2012 and 2015 Ageing Reports, the pension reforms anchored in the first two programmes are expected to deliver some positive results for the long-term sustainability of Greece’s pensions system. According to the programme conditionality, Greece should limit the increase in public-sector spending on pensions over the 2010-2060 period to under 2.5 percentage points of GDP. In light of the 2015 Ageing Report, pension reforms are expected to result in an overall decrease in public spending on pensions by 1.9 percentage points; a better performance than the programme target. However, they are projected to remain at the very high level of 14.3 % of the GDP in 2060. 68 Restoring financial stability 117. The extent to which financial sector reforms are successful depends also on factors external to the programmes. However, based on a range of indicators, it is possible to evaluate general developments in the banking sector throughout the programmes and thus establish how far their broad objectives have been achieved. The key aspects to be considered in this analysis are solvency, credit risk and asset quality, liquidity and profitability; these indicators were also in the spotlight of the programme conditions and the bank restructuring plans approved by the Commission. Solvency 118. Largely due to extensive programme funding (31.9 billion euros) and the sector’s restructuring, Greek banks’ capital ratios have improved considerably in line with international trends since the global crisis. Therefore, the main capital ratio (Common Equity Tier 1) was raised to 17 % by 2016, i.e. above the EU average of 14.2 % and the pre-crisis level (see Annex X). However, the high level of tax claims against the Greek state and nonperforming loans remained a concern for the banks’ solvency. Credit risk and asset quality 119. Over the programme period, the credit ratings of the four largest Greek banks declined significantly, with a drop to ‘default’ status in 2015 62. Despite completion of the third recapitalisation which strengthened banks’ capital adequacy, all bank ratings remained in the non-investment grade 63 range during 2016. 120. With regard to asset quality, Greek banks entered the global crisis in 2008 with outstanding NPLs totalling 5.5 % of the total loan portfolio (7 % in 2009), i.e. significantly higher than the euro area average. Due to the unprecedented economic recession, the 62 Moody’s: Caa3, Fitch: RD (restricted default), S&P: SD (selective default). In 2016, only S&P upgraded the four largest Greek banks to CCC+. 63 Non-investment grade (BB or lower) means that probability that the company will repay its issued debt is deemed to be speculative. 69 subsequent rise in unemployment and unaddressed strategic defaults by borrowers, NPLs have risen sharply throughout the programmes. In 2016 - based on a broader definition non-performing exposures reached the highest level in the EU (45.9 %) and were nine times higher than the EU average (5.1 %; see Annex X). Liquidity 121. Since the first programme, Greek banks have been asked to reduce their reliance on central bank borrowing (i.e. the Eurosystem) as their liquidity situation was already strained. In this context, the first programme did not succeed in reducing dependency on central bank borrowing, while the second and third programmes did. Nevertheless, in 2016 Greek banks continued to face tight liquidity conditions as their reliance on central bank funding remained well above the 2009 level (see Table 5); although it was lower than the mid-2015 peak (150 billion euros). As regards standard sources of liquidity, in 2016 there was no evidence of a sustained return of deposits (almost half in comparison with 2009, from 238 billion euros to 121 billion euros) and domestic banks still had limited access to other funding sources. Table 5 - Greek banks’ funding profile (in billion euros) 2009 2010 2011 2012 2013 2014 2015 2016 Central bank borrowing 49.7 97.7 128.7 121.2 73.0 56.0 107.6 66.6 Private-sector deposits 237.5 209.6 174.2 161.5 163.3 160.3 123.4 121.4 Source: Bank of Greece. Profitability 122. The programme policies that were put in place attempted to improve bank profitability through in-depth consolidation and restructuring. By considerably reducing staff numbers and branches, domestic banks significantly improved their cost-to-income ratio (51.9 %, compared to an EU average of 65.7 % as of 2016) and operating expenses. However, on a consolidated basis they remained loss-makers for most of the above period and their 70 performance in relation to a number of profitability indicators was still among the worst in the EU (see Annex X). 123. Overall, the programmes were unable to avert a sharp deterioration in Greek banks’ balance sheets, primarily due to adverse macroeconomic and political developments. The intermediary role of banks was undermined and the channels for financing the real economy were both restricted 64 and costly. Reviving growth 124. Following its integration with the Euro area in 2001, Greece enjoyed a period of very high economic growth, but suffered significant losses in competiveness at the same time. Correcting structural imbalances and barriers to competitiveness was therefore one of the programmes’ objectives, the aim being to restore sustainable growth in the long term and thus facilitate fiscal adjustment. This objective was not achieved, in the sense that Greece faced a protracted and exceptionally deep recession with serious consequences for the labour market. Key macroeconomic indicators performed systematically worse than the programmes had assumed (see Annex XI). Economic growth and its context 125. The Greek economy did not return to growth as quickly as other Euro area member states, including those which benefited from financial support programmes (e.g. Ireland and Portugal). Certain weaknesses in the programmes’ design and implementation (see Part II) were not conducive to facilitating recovery, but the macroeconomic situation was also due to a number of external factors and the depth of imbalances accumulated prior to the crisis. The programmes failed to meet their growth objectives; however, we do not know what the macroeconomic scenario would have been without the programmes. 126. The Greek economy consistently underperformed when compared with the growth objectives on which the macroeconomic projections were based. Between 2009 and 2016, 64 Outstanding loan amounts in the private sector fell by 21.9 % between 2009 and 2016. However, the actual negative flow of bank financing was much higher owing to the high and growing proportion of NPLs. 71 Greece’s economy shrank by over a quarter. Actual GDP growth in 2011 and 2012 was lower than the initial programme forecast (May 2010) by 6.5 p.p. and 8.4 p.p., respectively (see Figure 12). Although the forecast projected a return to positive growth in 2012, this actually occurred in 2014 and only for one year. The Greek economy also underperformed compared to the forecasts of March 2012 and April 2014, although the discrepancies were significantly smaller. Figure 12 - GDP growth in Greece and the Euro area (%) 4.0 2.1 2.0 2.1 1.1 0.4 0.0 -2.0 -4.0 -0.2 -2.0 0.0 -2.6 Greece 1st EAP projection -3.2 -4.0 2nd EAP projection -4.3 Euro area -5.5 -6.0 -7.3 -8.0 -9.1 -10.0 2009 2010 2011 2012 2013 2014 2015 2016 Source: EAPs, Eurostat. 127. During the programme period (up to 2015), Greece consistently achieved positive export growth, although this followed a deep export contraction in 2009. Compared to the programme forecast, exports generally underperformed, but the deviation was smaller than in the GDP forecast. In 2013, exports were at their 2008 level, whereas imports fell by 30 % over the same period which allowed Greece to achieve external re-balancing. 128. The recovery of exports as of 2010, however, cannot be fully attributed to the programmes’ reforms. Some relevant sectors such as tourism, benefited from internal 72 devaluation. However, sectoral analysis shows that Greek exports are also influenced by sectors not affected by domestic prices and wages 65 (e.g. oil products and maritime transport), whose performance reflect trends on global markets rather than an improvement in domestic competitiveness (see Annex XI). Prices and unemployment 129. Price levels did not develop in line with the programme’s projections. The initial forecast (May 2010) significantly underestimated inflation for 2010-2011 (also due to the impact of tax measures) and did not anticipate deflation in 2013-2014 (see Annex XI). Consequently, rising prices during the first programme slowed down the internal devaluation process which prevented Greece from improving its external competitiveness. Discrepancies between actual and forecast price developments persisted later in the programme, but became smaller. 130. In line with the deeper-than-expected economic downturn, the labour market significantly underperformed when compared with the initial programme assumptions (see Figure 13). Unemployment peaked at 27.5 % in 2013, even though it was initially projected to peak at 15.2 % in 2012. Unemployment also rose above the level projected in subsequent forecasts. The unemployment rate for the under-25s peaked at almost 60 % in 2013, but fell significantly to under 50 % in the following two years (see Annex XI). 65 With the notable exception of tourism, which benefited from internal devaluation. 73 Figure 13 - Unemployment level v. programme forecast (%) 30 27.5 24.5 25 24.9 17.9 20 15 10 26.5 Actual unemployment rate 12.7 1st EAP projection 9.6 2nd EAP projection 5 0 2009 2010 2011 2012 2013 2014 2015 Source: Eurostat and Greek Economic Adjustment Programmes. CONCLUSIONS AND RECOMMENDATIONS Management of the Economic Adjustment Programmes for Greece 131. The first Greek Economic Adjustment Programme was designed in a situation of extreme urgency and the management of all three programmes was impacted by political and economic uncertainties. The Commission was just one of several partners involved in the process as it was bound by the political guidance of the Eurogroup, and approval of the programmes required agreement between the Greek authorities and all Institutions representing lenders (the IMF, the Commission, the ECB, and the ESM under the third programme; see Introduction, paragraphs 22 to 23 and 31 to 32). 132. The Commission established programme management procedures almost one year after the launch of the first Greek programme. The procedures stipulated internal arrangements for information and approval flows at the Commission and Council and with programme partners, but provided no guidelines on the actual design of the programmes, for example in terms of the scope and level of detail of conditions (see paragraphs 24 to 25). 133. In the early stages of the programming process, some conditions were vague and did not specify any concrete actions. Conditions became clearer in the second programme, when their number and level of detail increased. Only in the third programme did the Commission mark important conditions as “key deliverables”, while in the first two 74 programmes the relative importance of the measures was not indicated (see paragraphs 26 to 28). Recommendation 1 The Commission should improve the general procedures for designing support programmes, in particular by outlining the scope of the analytical work needed to justify the content of the conditions and where possible by indicating the tools which could be drawn upon in relevant situations. In the case of future programmes, such guidance should make it easier for the Commission to organise its work in a situation of extreme urgency at the programme inception stage. Target date: end-2018 Recommendation 2 The Commission should prioritise conditions more effectively and specify the measures that are needed urgently in order to address the imbalances that are crucial for achieving the programmes’ objectives. Target date: immediately, if applicable 134. Despite the Commission’s efforts, programme conditions were not placed in the context of a broader growth strategy for Greece which could have lasted beyond the programmes. In the early stages the programmes’ immediate focus was on restoring fiscal sustainability, but the need for such a strategy became clearer over time. We found weaknesses in cross-policy coordination in programme design and cases where programme measures were not comprehensive in scope. Both impacted the effectiveness of structural reforms, in particular those that aimed to make the Greek economy more competitive (see paragraphs 29 to 30). 75 Recommendation 3 Where relevant to address the underlying economic imbalance, the Commission should ensure that the programmes are embedded in an overall growth strategy for the country, which is either already in place or has been devised with the Member State in the course of the programmes. If the strategy cannot be established at the programme inception stage, it should follow on as soon as possible and be reflected in any subsequent programme review. Target date: immediately, if applicable 135. Although the Commission did not have formalised procedures in place as regards the scope of monitoring, it regularly assessed and reported on compliance with programme conditions. Typically, the Commission was able to demonstrate a solid basis for its assessments, including some follow-up, such as detailed feedback on draft legal acts. Occasionally, we found problems in compliance monitoring, such as missing, inaccurate or vague assessments (see paragraphs 33 to 37). 136. Moreover, in the case of structural reforms, formal compliance did not always result in the changes actually envisaged by the programmes. Significant data gaps and a lack of adequate performance indicators for most of the structural conditions further impacted the relevance of the Commission’s monitoring in terms of being able to help understand the implementation and impact of reforms and take corrective actions when needed (see paragraphs 38 to 40). Recommendation 4 The Commission should have clear procedures and, where appropriate, KPIs to ensure that programme monitoring is both systematic and accurately documented. The monitoring of implementation, in particular for structural reforms, should focus more on effectiveness, go beyond the adoption of primary laws and also focus on the adoption of relevant secondary legislation and other implementing acts. The Commission should improve its arrangements for monitoring the implementation and roll-out of reforms so as to identify better administrative or other impediments to the effective implementation of the reforms. The Commission needs to ensure that it has the necessary resources to undertake such assessments. Target date: end-2018 76 Recommendation 5 The Commission should address data gaps more comprehensively from the outset of the programmes. It should also clearly specify all the data it needs to monitor the programmes and their results. Target date: immediately, if applicable 137. Despite complex institutional arrangements for the programmes, the internal working arrangements in terms of how the Commission cooperates with programme partners, primarily the IMF, ECB and ESM, have never been formalised. There was inadequate transparency regarding the partners’ roles, the division of responsibilities or working methods. However, the institutions have informally established effective arrangements for cooperation and information sharing (see paragraphs 41 to 43). Recommendation 6 The Commission should seek to reach an agreement with programme partners’ clarifying their roles and cooperation methods which should be transparent and sufficiently detailed. Target date: end-2018 138. The overall programme design was based on macroeconomic projections and funding gap calculations that had been prepared for each review by the Commission in cooperation with programme partners. The Commission’s analysis in this respect was internally consistent and based on the most recent data. However, we found a number of methodological weaknesses, in particular in the documentation of assumptions, the justification of judgemental inputs and the limited use of sensitivity testing. There were no formalised quality arrangements to ensure that calculations were accurate. Macroeconomic projections largely underestimated the economic crisis in Greece, but the errors were generally no larger than those of other international organisations (see paragraphs 44 to 54). 77 Recommendation 7 The Commission should better justify the assumptions for and modifications to the economic calculations underlying the programme’s design, including through appropriate publications. Such processes should be subject to appropriate quality control. This should apply in particular to programme reviews, which are carried out under less urgent circumstances than at the programme inception stage. Target date: end-2018 Design and implementation of the reforms 139. Financial support for Greece was conditional upon the implementation of a broad range of reforms addressing fiscal, financial and structural imbalances. The scope of the reforms evolved, but from the outset covered almost all functions of the Greek state, meaning in some cases that very deep structural imbalances had to be corrected. The Greek reform agenda was particularly challenging and its implementation was impacted by the political instability (see paragraphs 55 to 57). 140. Tax policy reforms delivered fiscal results, which mostly relied on changes in tax rates or new taxes. This contributed to instability in tax policy, which occurred for a variety of reasons including the lack of sustained reform efforts, and this ran counter to earlier assessments that prioritised broadening the tax base as a way of increasing tax revenue. The Commission did not sufficiently justify some key measures (e.g. VAT and excise-duty hikes) by providing an analysis of alternative options and their consequences. Tax administration reforms brought less tangible results, attributable to problems with implementation capacity, but also to design weaknesses such as late inclusion in the programmes and limited scope (see paragraphs 58 to 66). 141. The simultaneous implementation of fiscal savings, justified for the reasons of fiscal consolidation, as well as long-term structural changes was a challenge for public administration reforms, with the latter proving less successful. We found that the main causes on the design side were the insufficient consideration given to the administrative capacity to implement the reforms, and weak linkages between the programmes and other instruments, such as technical assistance and operational programmes under the structural 78 funds. The Commission could not justify key quantitative targets for the public administration reforms that were anchored in the programmes (see paragraphs 67 to 76). 142. Financial reforms ensured short-term stability in the sector, but a number of structural weaknesses were not comprehensively addressed or were included late in the programme (e.g. the management of Non-Performing Loans and corporate governance). The measures were more comprehensive under the third programme. We also found weaknesses in the risk analysis underpinning the bank recapitalisation of 2013 and inefficiencies in the design of subsequent recapitalisations. The sequencing of certain reform aspects was not sufficiently coordinated (see paragraphs 77 to 94). 143. The Commission was able to demonstrate sound analysis behind the labour market reforms which brought tangible results in terms of market liberalisation. However, implementation was hampered by significant delays. Some conditions, in particular at the programme inception stage, were not sufficiently geared to the structure of the Greek economy, and other key measures were included in the programme after some delay (see paragraphs 95 to 102). Recommendation 8 In order to mitigate potential weaknesses, the Commission should be more systematic when assessing the member states’ administrative capacity to implement reforms. Technical assistance needs and possible support from the Commission's Structural Reform Support department should be assessed in a programme’s early stages in cooperation with the MS’s authorities. The choice, level of detail and timing of conditions should be geared to the results of the analysis. Target date: immediate, if applicable 79 Recommendation 9 The Commission should enhance its analytical work on design of the reforms. It should, in particular, address the appropriateness and timing of measures, given the specific situation in the Member State. Target date: immediate, if applicable Achievement of programme objectives 144. By participating in the three Economic Adjustment Programmes, Greece avoided default. However, the achievement of the programmes’ objectives – fiscal sustainability, financial stability and a return to growth – was successful only to a limited extent. An external evaluation could have been useful to comprehensively understand the underlying complex developments, but the Commission has not carried one out (see paragraphs 103 to 106). 145. In the case of fiscal sustainability, the programmes brought about a significant consolidation: the primary structural balance was adjusted by 17 % of GDP in 2009-2015. However, in the same period, GDP shrank by over a quarter and unemployment exceeded 25 %. Only in 2014 did Greece record economic growth, although the programmes’ initial assumption projected a sustainable return to growth as of 2012 (see paragraphs 107; 111 to 113; 124 to 130). 146. Poor macroeconomic performance, coupled with financing costs on previously accumulated debt, mean that Greece has been consistently increasing its debt-to-GDP ratio throughout the programme period (except in 2012 due to the PSI). It was also unable to finance its needs on the markets. In the immediate post-programme period, Greece will have to repay substantial amounts of debt and the programme’s assumption is that they will be fully funded from the primary surplus and market financing. On the financial side, the programmes ensured the short-term stability of the financial system, but were unable to avert a sharp deterioration of the banks’ balance sheets primarily due to adverse macroeconomic and political developments (see paragraphs 114 to 123; 108 to 110). 80 Recommendation 10 The programmes should be subject to ex-post evaluation at least after they have expired. In the case of successive programmes the combined duration of which is substantially longer than the standard period of three years, an interim evaluation should be carried out and the results used to assess their design and monitoring arrangements. Target date: end-2018 Recommendation 11 The Commission should analyse the appropriate support and surveillance framework for the period after the programme ends. This should be done sufficiently in advance of the end of the programme and should take into account the financing needs of the country. Target date: immediate This Report was adopted by Chamber IV, headed by Mr Tomé MUGURUZA, Member of the Court of Auditors, in Luxembourg at its meeting of 3 October 2017. For the Court of Auditors Klaus-Heiner LEHNE President Annex I THE CHRONOLOGY OF THE GREEK ECONOMIC CRISIS Pre-Programme 2009 25 May 2009 IMF staff states that that the Greek banking system appears to have enough capital and liquidity buffers to weather the expected slowdown of the economy. 4 October 2009 PASOK wins the Greek parliamentary elections. The party obtains 43.92 % of the popular vote and 160 out of 300 parliamentary seats. 20 October 2009 The new finance minister discloses that the nation’s deficit will rise to almost 12.5 % of GDP. 22 October 2009 Fitch rating agency downgrades Greece's credit rating to A- from A. The downgrade is a result of the likelihood of Greece’s fiscal deficit reaching 12.5 % of GDP in 2009 8 December 2009 Fitch rating agency downgrades Greece's credit rating to BBB+ from A-. The downgrade reflects concerns over the medium-term outlook for public finances. Fitch's assessment is that the government debt burden is likely to rise to close to 130 % of GDP before stabilising. 16 December 2009 Standard and Poor's rating agency downgrades Greece’s credit rating. 22 December 2009 Greek parliament passes the 2010 budget. 22 December 2009 Moody’s rating agency downgrades Greece to category A2 from A1. The downgrade reflects very limited short-term liquidity risks and medium- to long-term solvency risks. 2010 21 January 2010 Greek/German 10-year debt yield spread exceeds 300 basis points. 9 February 2010 Parliament approves the first fiscal measures, including a freeze in the salaries of all government employees, a 10 % cut in bonuses, and cuts in overtime. 3 March 2010 Parliament passes a major new fiscal package. The measures include: Pension freezes; an increase in VAT standard rate from 19 % to 21 % and of excise duties on fuel, cigarettes, alcohol and luxury goods; and cuts in public-sector pay. 9 April 2010 Fitch downgrades Greece’s credit rating to BBB- from BBB+. 12 April 2010 Euro-area finance ministers agree to provide Greece with up to 30 billion euro in loans over the coming year, with the IMF agreeing to put up another 15 billion euro in funds. 23 April 2010 Greece’s prime minister formally requests international support. The law 3842 reforms personal income tax and adopts anti-tax evasion measures. 27 April 2010 Standard and Poor's downgrades Greece's credit rating below investment grade to junk bond status. 28 April 2010 Greek/German 10-year debt yield spread exceeds 1000 basis points. st 1 Adjustment Programme 2 May 2010 The Greek Prime Minister, the IMF and euro-zone leaders agree to a 110 billion euro bailout package that will take effect over the following three years. The government announces the new fiscal package measures. 2 6 May 2010 The Greek parliament passes the new fiscal package measures. The bill passes with 172 votes in favour and 121 against. The VAT standard rate is further increased from 21 % to 23 % as from 1 July 2010. 7 July 2010 Parliament passes pensions reform, a key EU/IMF requirement. 13 July 2010 Creation of the Hellenic Financial Stability Fund with the aim of contributing to the financial stability of the Greek banking system. 4 October 2010 Greece announces a draft budget plan to cut the deficit to 7 % of GDP in 2011. 15 December 2010 Parliament passes the law on public companies, capping on monthly wages and cutting salaries above 1800 euro by 10 %. 23 December 2010 Greece’s parliament approves the 2011 budget. 2011 14 January 2011 Fitch downgrades Greece’s credit rating to BB+ from BBB-. 7 March 2011 Moody’s downgrades Greece’s credit rating to B1 from Ba1. 11 March 2011 The EU reaches a preliminary agreement to cut rates on emergency loans to Greece by 100 basis points for first three years and extend loan maturities to 7.5 years. 29 March 2011 A new law adopts measures against tax evasion. The corporate income tax rate is reduced from 24 % to 20 %. 20 May 2011 Fitch downgrades Greece’s credit rating to B+ from BB+. The rating downgrade reflects the challenge facing Greece in implementing a fiscal and structural reform programme, and the political risk of implementing further fiscal measures. 1 June 2011 Moody's downgrades Greece to default category (Caa1) from B1 because Greece is ‘increasingly likely to fail to stabilize its debt ratios within the timeframe set by previously announced fiscal consolidation plans’. 13 June 2011 Standard and Poor’s downgrades Greece to default category too. 29 June 2011 The Greek parliament passes the new fiscal measures despite public demonstrations. The bill passes with 155 votes in favour and 138 against. The measures include new taxes (e.g. a new tax for individuals whose annual income exceeds 12 000 euro), and new cuts in workers’ wages. 13 July 2011 Fitch downgrades Greece’s credit rating to CCC (default category) from B+. 15 July 2011 EBA publishes the results of its EU-wide stress testing exercise. 25 July 2011 Moody’s downgrades Greece’s credit rating to category Ca-. 8 August 2011 The general stock index falls below 1000 points, its lowest level since January 1997. 24 August 2011 Bank of Greece activates emergency liquidity assistance (ELA) mechanism to support Greek banks. 2 September 2011 Institutions’ inspectors suspend Greece’s fifth review after finding delays in implementing the medium-term fiscal plan and structural economic reforms. 3 27 September 2011 The Parliament adopts new taxation measures (a broad recurrent property tax, the personal income tax free threshold is decreased from 12 000 euros to 5000 euros). 20 October 2011 Greece’s government passes a comprehensive savings bill amid protests and riots outside the parliament building. 27 October 2011 Investors agree to a "haircut" of 50 % by converting their existing bonds into new loans. 31 October 2011 Greece’s Prime Minister calls a confidence vote and a referendum to approve the EU summit deal about the Greek debt haircut. 6 November 2011 The Prime Minister resigns. 10 November 2011 Lucas Papademos becomes Greece’s new Prime Minister and head of the three-party coalition government. 2012 12 February 2012 Parliament passes a new round of fiscal cuts amid protests. Many buildings in the centre of Athens are torched during the riots. nd 2 Adjustment Programme 9 March 2012 Restructuring of the Greek sovereign debt under the PSI is completed successfully. 14 March 2012 Euro-area finance ministers reach agreement on a second bailout package for Greece. The deal includes a 53.5 % write-down for investors in Greek bonds. 6 May 2012 Election held. The New Democracy party wins, but with a lower share of the vote. As no party wins a majority, a new election is called for early June. 16 May 2012 Panagiotis Pikramenos becomes interim Prime Minister. 17 June 2012 Early election held. New Democracy wins with 29.7 % of the vote, but does not win a majority in parliament. Four days later, it forms a coalition government with New Democracy, PASOK and DIMAR. Antonis Samaras, the leader of New Democracy, becomes the new Prime Minister of Greece. 27 July 2012 Agricultural Bank of Greece is put into resolution and Piraeus Bank acquires all its sound assets. 5 November 2012 The Greek parliament adopts a new round of fiscal measures that are required for Greece to receive the next instalment of the international economic bailout. The social security contributions are reduced by 1.1 percentage point. 11 November 2012 Greece passes the 2013 budget. 2013 11 January 2013 The Greek parliament adopts new taxation measures (increase of corporate income tax rate from 20 % to 26 %, a new reform of the personal income tax with an increased tax-free threshold, elimination of tax free bracket for the self-employed, replacement of the children tax credits with means-tested benefits etc.). 1 February 2013 Alpha Bank acquires Emporiki Bank from Credit Agricole 4 25 March 2013 Eurogroup reaches an agreement with Cyprus, which includes an agreement between Cyprus and Greece on the transfer of Greek branches of Cypriot banks to Piraeus Bank. 28 April 2013 Parliament approves a bill cutting 15 000 state jobs by the end of the following year, including 4 000 in 2013. 24 June 2013 Greek Prime Minister Antonis Samaras reshuffles his cabinet. 17 July 2013 The Greek parliament approves new austerity measures, including a plan for thousands of layoffs and wage cuts for civil servants. 2014 6 March 2014 Bank of Greece indicates the capital requirements (6.4 billion euro for 2013-2016 on the basis of the baseline scenario) for the Greek banking sector. 7 April 2014 The Greek parliament further reduces the social security contributions by 3.9 percentage points. 26 October 2014 ECB publishes the results of its comprehensive assessment of 130 European banks. 19 December 2014 The EFSF programme for Greece is extended by two months (until 28 February 2015). 30 December 2014 Early elections called for 25 January 2015 after inability to elect a new head of state. 2015 25 January 2015 Early elections result in victory for SYRIZA, which wins 36.34 % of the vote and 149 out of 300 seats. 27 January 2015 Formation of coalition government. 4 February 20151 ECB suspends the waiver for the eligibility of Greek bonds as loan collateral. 27 February 2015 New extension of the EFSF programme for Greece by four months (until 30 June 2015). 18 June 2015 Eurogroup President announces that Greece's progress is insufficient and a new agreement is needed. 22 June 2015 Greek government sends a new proposal to the Eurogroup. 26 June 2015 Greek referendum called for 5 July 2015. 28 June 2015 Greece closes its banks and imposes capital controls. 30 June 2015 Greece fails to repay 1.5 billion euro to IMF. 5 July 2015 61.3 % of Greeks vote “No” to the draft agreement submitted by the institutions at the Eurogroup meeting of 25 June 2015. 6 July 2015 Finance Minister resigns. New Minister Euclid Tsakalotos appointed. Bank holiday extended. 8 July 2015 Greece submits an official request for a third assistance programme – in the form of a loan facility – to the ESM to be used to meet debt obligations and ensure the stability of its financial system. 5 13 July 2015 IMF statement on Greece’s payments due (456 million euro). Greek authorities request an extension. 15 July 2015 The Greek parliament adopts new tax measures (the corporate income tax rate is increased from 26 % to 29 %; the rates of the solidarity contribution for individuals’ annual incomes higher than 30 000 euros are increased; the scope of VAT standard rate is extended). Several ministers resign. 16 July 2015 Eurogroup grants 3-year ESM stability support to Greece. 17 July 2015 Government reshuffle to replace ministers who rejected the loan agreement. Council approves 7 billion euro bridging loan to Greece. Bank holiday ends. 20 July 2015 Debt repayment to IMF, ECB and Bank of Greece. 19 August 2015 Memorandum of Understanding detailing the economic reform measures and commitments associated with the financial assistance package. rd 3 Adjustment Programme 14 August 2015 Parliament approves the third bailout programme with the support of five out of seven political parties. ESM proposal for the terms of the first tranche of 26 billion euro under the Financial Assistance Facility Agreement for Greece. 19 August 2015 ESM Board of Governors approves ESM programme for Greece - Memorandum of Understanding (MoU) detailing the economic reform measures and commitments associated with the financial assistance package. 20 August 2015 Prime Minister Alexis Tsipras announces the resignation of his government and calls early elections for 20 September 2015. 27 August 2015 The Head of the Supreme Court sworn in as interim Prime Minister. 20 September 2015 Greek legislative election, SYRIZA-ANEL coalition government formed. 7 October 2015 Government receives vote of confidence from Parliament with 155 votes in favour and 144 against. 31 October 2015 After a stress test ECB founds a capital shortfall of 14.4 billion euro in the four largest Greek banks. BoG also founds a capital shortfall of 0.7 billion euro in the largest less significant bank. 19 November 2015 In order to receive a disbursement of 2 billion euro, Parliament approves prior actions with a small majority of 153 votes. 16 December 2015 Introduction of Law No. 4354/2015 for NPLs sales and servicing. 2016 8 May 2016 New pensions reform – new regulations on income taxes and social security contributions. 22 May 2016 Further increases of various indirect taxes rates. VAT standard rate is increased from 23 % to 24 %. 23 May 2016 Preliminary Debt Sustainability Analysis of Greece by the IMF. 6 9 June 2016 First review of the third Economic Adjustment Programme is completed – compliance report issued. 16 June 2016 Agreement on the first update of the Memorandum of Understanding with accompanying Technical Memorandum of Understanding. 22 June 2016 ECB reinstates the waiver for the eligibility of Greek bonds. 10 December 2016 Vote on the 2017 budget: 152 votes in favour and 146 against. 7 Annex II CAPITAL CONTROLS Greece has been the second economy, following Cyprus in late March 2013, to introduce restrictions on 28 June 2015 in the free movement of capital in the context of a monetary union. Unlike several previous incidences, which mainly aimed to support exchange rate stability (Iceland 2008, Brazil 2009 and India 2013), capital controls in Greece were imposed to curtail deposit outflows and safeguard domestic financial stability, in a period of high uncertainty regarding political developments and domestic macroeconomic prospects. After the failure of negotiations on the closure of the second programme and the announcement of the referendum, ECB decided to cease raising the overall emergency liquidity assistance (ELA) ceiling for Greek banks that had observed a ‘bank-walk’ since October 2014 (with outflows of approximately 42 billion euro or 26 % of private sector’s deposits in eight months), prompting the Greek authorities to impose a bank holiday, and subsequently, capital controls on deposits and transactions abroad. Greek banks remained closed for a period of three weeks. During this period, customers were not allowed to withdraw more than 60 euro per bank card and per day. All transactions to foreign banks had to get prior approval from a government body. However, there was no limit on domestic transactions by debit and credit cards, and on withdrawal with cards issued outside Greece. On 12 July 2015, the Euro Summit agreement paved the way for a lift in the ELA ceiling and the subsequent reopening of Greek banks. A number of restrictions have been gradually relaxed, with the latest changes taking place in July 2016. The imposition of capital controls led to an immediate tightening of the liquidity situation for Greek households and businesses alike, which made the Greek economy return to recession, disrupting five consecutive quarters of positive growth. However, the contraction of aggregate economic activity has proven to be milder than initially expected by the institutions (2015 real GDP: -0.2 %). A number of domestic and external drivers (e.g. the agreement on the third programme, good tourism season, completion of bank recapitalisation, drop in oil prices and devaluation of euro) had contributed in the milder recession. Private sector deposits also largely stabilised. However, a faster return of deposits would have required a further significant recovery of economic sentiment. Source: European Parliament’s EGOV, Greece’s financial assistance programme, March 2017; Eurobank, One year capital controls in Greece: Impact on the domestic economy and lessons from the Cypriot experience, 1 August 2016. 8 Annex III WEAKNESSES IN COMPLIANCE MONITORING Taxation Condition MoU st Monitoring “Parliament adopts legislation to improve the efficiency of the tax administration and controls, implementing recommendations provided by the European Commission and IMF. In particular, they put in place an effective project management arrangement (including tight MOF oversight and taskforces) to implement the anti-evasion plan to restore tax discipline …” 1 Programme, initial MoU The condition was flagged as “partly observed” together with another condition (2nd review of 1st programme) but it is unclear what is not observed (e.g. no assessment on adopted laws, on task forces). The justification of the assessment refers to facts that are not part of this condition. “Restore tax discipline through: … a reorganized large taxpayer unit focused on the compliance of the largest revenue contributors;” 1st Programme, initial MoU The condition was flagged as “observed” (2nd review of 1st programme) while in reality it was not: new conditions of the revised MoU required its implementation. The condition was not assessed for compliance (it was reported as “N/A” in 4th review of 2nd programme). “The Government continues work on a standard 2nd Programme, procedure for revision of legal values of real estate 3rd review to better align them with market prices that will be in place for the purposes of capital taxation for the fiscal year 2016, and issues a status report on the work and a detailed timetable (September 2013).” 2nd Programme, Initial and 1st reviews 2nd Programme, 4th review Not assessed for compliance “To clear expenditure arrears and tax refunds, the conditions for a government unit to meet to allow funds for clearance to be disbursed will include, for expenditure arrears: (i) establishment by the unit of a fully functioning commitment register and (ii) reporting of at least three months of consistent data on commitments, payments, and arrears (2 months for EOPYY); and, for both expenditure arrears and tax refunds: (iii) verification of claims.” 2nd Programme, 2nd review No compliance assessment at deadline (2nd review of 2nd programme) and not followed up after that. Separately, a new law on taxation of real estate will be adopted by end-June for 2014 onwards. The new real estate taxation regime, which will consolidate a number of now separate taxes, will be designed to be budget neutral ensuring annual revenue of at least 2.7 billion euro. 2nd Programme, 2nd review This prior action was reported as “observed” in the 2nd review of 1st programme but it was not in reality, as the unified property tax (ENFIA) was adopted by law (no 4223) in December 2013, based on an The prior actions in taxation area. 68 conditions in taxation area. Not assessed for compliance 9 The Government will prepare a tax reform that aims at simplifying the tax system, eliminating exemptions and preferential regimes, thus broadening bases, and allowing a gradual reduction in tax rates as revenue performance improves. This reform relates to the personal income tax and corporate income tax. The reform will be adopted in December 2012 to entry into force in 2013 2nd Programme, 1st review “The Government will: adopt legislation to introduce a new Income Tax Code that will simplify the existing law, increase its transparency, and remove ambiguities, whilst allowing easier administration, encouraging tax compliance, and ensuring more robust revenue through the cycle. The new income tax code will reduce filing requirements for payas-you-earn taxpayers and those who receive investment income, consolidate cross-border merger and reorganization provisions, and introduce anti-avoidance provisions to combat international tax avoidance;” “The Authorities: ii. present a plan to replace payments in cash and cheque in tax offices with bank transfers (July 2013);” (third MOU update, second programme) 2nd Programme, 3rd review additional condition of the third MoU update: “The Government will pass legislation on the property tax regime to take effect in 2014.” This condition was flagged as “observed” in the 2nd review of 2nd programme while it was not (the law 4110 of January 2013 contained only parametric changes to the personal income tax).A similar condition was added in the next (second and third) MoU updates “The Authorities will adopt a unified and simplified income tax code”. The new income tax code was voted in July 2013 as a prior action of the third MoU update (law 4172/2013). The condition was considered as “observed” in the 3rd review (law 4172/2013 was adopted). However, the law was not operationalised because the authorities needed time to adopt the implementing rules. 2nd Programme 4th review While it was due in July 2013, it was assessed as “complied” with a ninemonth delay in April 2014 (4th review of 2nd programme). Condition MoU Monitoring Following dialogue with social partners, government adopts legislation on minimum wages to introduce sub minima for groups at risk such as the young and long-term unemployed, and put measures in place to guarantee that current minimum wages remain fixed in nominal terms for three years. 1st Programme, Initial MoU The requirement of introducing subminima for young employees was indeed “complied with”. However, there is no evidence of the three year freeze of minimum wage and introduction of sub-minima for longterm unemployed. At the first review in August 2010 indeed the conditions were not yet due, but they were also not repeated and assessed in the subsequent MoU. In 2014 (Law 4254/2014 of 8 April) minimum wage top-ups for longterm unemployed have been reduced, yet this measure cannot be Labour 10 considered as adequate with the programme’s requirement of introducing “sub-minima”. The compliance status is “partially observed”, although the use of special firm-level collective agreements was marginal and data incomplete (see para 76). Further, no report was produced (the only actual “output” required in this MoU condition) and we received no evidence that it was done for the subsequent reviews. Government promotes, monitors and assesses the implementation of the new special firm-level collective agreements. It ensures that there is no formal or effective impediment to these agreements and that they contribute to improve the adaptability of firms to market conditions, with a view to create and preserve jobs and improve the firms’ competitiveness, by aligning wage developments with productivity developments at firm level. It provides a report on its assessment. Any other amendment to the law on sectoral collective bargaining is adopted before end-July 2011. 1st Programme, 5th review Government reviews the current structure of the minimum wage rates system, with a view to possibly improve its simplicity and effectiveness to promote employability and fight unemployment and enhance the competitiveness of the economy. 2nd Programme, 4th review Reported as “not-observed” under 4th review of 2nd programme; implemented in April 2014, but was never subjected to formal compliance assessment. MoU Monitoring Business environment Export Facilitation Condition st Government takes measures, in line with EU competition rules, for the adoption of measures to facilitate PPPs Government takes measures, in line with EU competition rules, to strengthen export promotion policy. 1 Programme, 1st review Government carries out in depth evaluation of all R&D and innovation actions, including in various Operational Programmes, in order to adjust the national strategy. 1st Programme, 3rd review Government creates an external advisory council financed through the 7th R&D programme, to consider how to foster innovation, how to strengthen links between public research and 1st Programme, 3rd review 1st Programme, 3rd review No compliance assessment. Conditionality repeated in the 1st review, then discontinued. The compliance status is ambiguous since it is not clear if the status “observed” applies only to the adoption of the investment law or includes also the export promotion policy. Moreover, export promotion appears again in the 3rd EAP with a deadline sets at December 2015. In the supplemental MoU of June 2016, the condition is postponed again to May 2016. Condition not complied with. The compliance status is ambiguous. The review states that the condition is “observed” but asks for the finalisation of the creation of an 11 Greek industries and the development of regional industrial clusters. external advisor council. The fourth programme review mentioned “not observed” for a set of three conditions including this one. Nevertheless, the last part of the sentence (i.e. “according to government”) suggests an inadequate assessment. Financial sector reforms Condition MoU Monitoring Review the private sector bankruptcy law to ensure consistency with ECB observations. 1st Programme, 1st review The Bank of Greece (BoG) and the Government ensure that the Hellenic Financial Stability Fund (HFSF) is fully operational. 1st Programme, 2nd review The BoG commits to reduce remuneration of its staff in light of the overall effort of fiscal consolidation. 1st Programme, 2nd review Following up on the result of the July 2010 Committee of European Banking Supervisors (CEBS) stress tests, the bank which did not pass the test implements interim restructuring measures under enhanced supervision by the BoG. Government provides its full support to this bank and ensures that it complies with the requirement of implementing a restructuring plan under the EU rules for state aid, including compliance with the 1.10.2010 deadline for submission. Government tables legislation that places all registered banks' employees under the same private sector status, regardless of the bank ownership. 1st Programme, 2nd review The condition was assessed as “ongoing” in the 1st review, but no follow up has been made after that and the compliance assessment remains unclear. The condition was flagged as “partially observed” at deadline in the 2nd review. The last assessment performed at the 5th review, staffing issues were noted. No further assessment was made after that, thus the compliance status remains unclear. The condition was assessed as “partially observed” in the 2nd and 5th reviews of 1st MoU. We could track this condition again to the 2nd MoU but it was not further followed up. Therefore, the compliance status remains unclear. The condition was assessed as “partially observed / ongoing” in the 2nd review. The condition was then not followed up in the 3rd review of the 1st MoU even though it had been described as only “partially observed” and the assessment was still ongoing. 1st Programme, 4th review The condition was not observed at the 4th and 5th review of the first MoU. No follow-up has been performed after this date. Compliance status is therefore unclear. 12 To support banks in their effort to restructure operations, Government takes steps to limit bonuses and to eliminate the so- called “balancesheet premium,” or other equivalent measures. 1st Programme, 5th review The condition was not observed in the 4th and 5th reviews; furthermore no follow-up has been performed. Compliance status is therefore unclear. The BoG transfers staff with prerequisite specialist skills into the supervision department. It will also consider requesting long-term technical assistance to be resourced from other European supervisory authorities. 1st Programme, 5th review The BoG will require the banks whose capital base has fallen short of regulatory requirements to take appropriate actions via capital injections or restructuring. Should it include the involvement of other local financial institutions, the BoG will submit a financial impact analysis and legal opinion. 1st Programme, 5th review The condition was flagged as “partially observed” at the 5th review, but it was then not followed up in the 2nd MoU. Compliance status is therefore unclear. The condition was flagged as “partially observed” at the 5th review. Due to the length and complexity of the procedures, the agreed deadline was not attainable. However, condition was not followed up in the 2nd MoU. Public administration reforms Condition MoU Monitoring Government adopts an act that limits recruitment in the whole general government to a rule of not more than 1 recruitment for 5 exits, without sectoral exceptions. 1st Programme, 2nd review Although the condition was assessed as “complied”, it included multiple sub-conditions that were not complied with. The programme assessment does not sufficiently cover all sub conditions. Law 3899/2010 establishes that government recruitments in the course of 2011-13 need to fit in the rule of not more than one recruitment for every five exits. However, there is no comprehensive human resource plan yet. Also, the absence of regular and timely compilation of staff movements (entries, exits and transfers) means that the enforcement of the rule cannot be monitored. This condition included multiple sub-conditions. The programme assessment does not sufficiently cover all sub conditions. The draft report includes a diagnosis on wages and employment data in the public sector. Nevertheless, it does not contain plans for the allocation of human resources in the public sector for the period up to 2013. Government prepares a human resource plan in line with this rule. The rule also applies to staff transferred from public enterprises under restructuring to government entities. The Ministry of Finance together with the Ministry of Interior complete the establishment of a Single Payment Authority for the payment of wages in the public sector. The Ministry of Finance prepares a report (to be published by end January 2011), in collaboration with the Single Payment Authority, on the structure and levels of remuneration and the volume and dynamics of employment in the 1st Programme, 2nd review 13 general government. The report presents plans for the allocation of human resources in the public sector for the period up to 2013. It specifies plans to reallocate qualified staff to the tax administration, GAO, the labour inspectorate, regulators and Hellenic Competition Commission. Based on a sample of conditions subject to the audit. 14 Annex IV TAXATION REFORMS Part A: Data gaps Field Comment Property taxation Tax revenue data were not sufficiently detailed during the first programme: the Commission did not point out any problem with the collection of property tax on real estate “FAP” in the monitoring reports even though there was basically no collection in 2011-2012. Personal income taxation The specification of income tax reforms required detailed data sets on the distribution of personal income, but these were not available during the first programme. Tax refund claims The monitoring of tax refund claims was first requested in the first update of the memorandum of economic and financial policies “MEFP” for the second programme. It is unclear why the data were not requested during the first programme, even though unpaid tax refund claims are usually a risk in assistance programmes. During the second programme, there were regular inconsistencies between changes in stocks and flows of the reported claims. Tax evasion The programmes did not provide for regular collection of the data needed to estimate tax evasion or for any monitoring of tax gaps for different taxes and economic sectors, which were flagged as relatively high in the pre-crisis period. There were no monitored estimates of the impact of undeclared work on personal income tax and social security contributions revenue. Similarly, there was no analysis of data on the Labour Inspectorate’s performance in the area of undeclared work Tax debts Tax debts were monitored as from the second programme, but the indicators were not clearly defined from the outset. The data and targets referred only to the collection of tax debts, while the tax debt stock was not defined, monitored or targeted. This was inconsistent with the fact that tax debts persistently increased over the course of the two programmes. In addition, social security debt was not monitored during the programmes. Performance indicators for tax administration No performance indicator for tax administration was established during the first programme. Output indicators (such as the number of tax inspections) were put in place during the second programme, but only for tax administration, and therefore excluded social security funds. The OECD published indicators for the efficiency of tax administration (‘Tax administration 2015’) but data were not available for Greece in this report (e.g. expenditure and staffing-related ratios, the cost of collection ratios, the ratio of salary costs to administrative costs, staff turnover/attrition rate etc.). No indicators were monitored/collected by the Commission for tax administration or social security funds during the programmes. The key performance indicators put in place by the tax authorities with the support of the second programme were not aligned with these indicators. Part B: Instability in tax policy A transparent, simple and stable tax system is generally considered to be an important means of encouraging investment. Instability in tax policy affected all main types of taxes in Greece: changes and reversals in the scope of VAT rates (see Table 1), changes in corporate income tax rates, property and investment taxation, and the taxation of employed and self-employed labour. In particular, the corporate income tax rate was reduced from 25 % to 24 % in 2010 then to 20 % in 2011 and was 15 increased to 26 % in 2013 and to 29 % in 2015. OECD and Commission data also show that the tax wedge for a single employee increased by 2012 and returned to the 2009 level in 2015 with unwarranted economic objectives for the labour market. In particular, the social contribution rates and the threshold for tax-exempted personal income was raised and then lowered. Instability in tax policy was also generated by multiple reforms of taxable income determination, bookkeeping, tax audits, penalties etc. For example, the first programme advocated unified tax treatment of personal income sources but the second programme moved in the opposite direction. The tax procedures and the Code of Books and Records were both subject to changes during the first and second programmes. The former income tax code (Law 2238/1994) was amended 425 times by 34 laws in the course of the programmes (2010-2014). When it was adopted in 2013, the new income tax code (Law 4172/2013) did not abolish the previous tax code, a situation which created temporary legal uncertainties for taxpayers. The adoption of the new income tax code was a requirement of the MoU. It was also amended 111 times by 20 laws during 2013-2014. In the ‘Global Competitiveness Report 2015-2016’, Greece was ranked 136 out of 140 as regards the effect of taxation on investment incentives. Table 1 - Main VAT rate changes (2010-2015) Enforcement date 15 March 2010 Change VAT rates were increased from 19 %, 9 % and 4.5 % to 21 %, 10 % and 5 %. The rates for specific islands were increased from 3 %, 6 % and 13 % to 4 %, 7 % and 15 %. 1 July 2010 VAT rates were increased to 23 %, 11 % and 5.5 %. Exonerated legal, private health and cultural services provided by the private sector were brought under the standard rate. The rates for specific islands were increased to 4 %, 8 % and 16 %. 1 January 2011 Reduced rates were increased again to 13 % and 6.5 %. The rate for hotel accommodation and pharmaceuticals changed from 13 % to 6.5 % (this was a setback for the objective of broadening the base of the standard rate). 1 September 2011 The rate for non-alcoholic beverages, restaurants/cafés/take-away/ready-made food in supermarkets increased from 13 % to 23 %. 1 August 2013 The rate for non-alcoholic beverages, restaurants/cafés/take-away/ready-made food in supermarkets decreased from 23 % to 13 % (this was a reversal of a previous measure). 16 20 July 2015 The rate for transportation of passengers, non-alcoholic beverages, restaurants/cafés/take-away/ready-made food in supermarkets increased from 13 % to 23 % (this included a reversal of the previous measure). 1 October 2015 The rate for hotel accommodation increased from 6.5 % to 13 % (this was a reversal of a previous measure). The rates for certain islands increased to mainland rates. Part C: Examples of unrealistic deadlines Condition Comment ‘The Government discontinues payments in cash and by cheque in tax offices and these should be replaced by bank transfers so that staff time is freed up to focus on more valueadded work (audit, collection enforcement and taxpayer advice); [Q2-2012]’ (initial MoU, second programme). The IT project could not realistically meet a target of three months. The deadline was extended in the first MoU update (to December 2012) and the second MoU update (to June 2013). In the third MoU update, the condition was changed into a requirement to draft a plan for deploying the necessary electronic means. By the end of 2014, the project had not been completed. ‘The Government continues to centralise and merge tax offices; 200 local tax offices, identified as inefficient, will be closed, by end2012’ (initial MoU, second programme). The deadline was extended several times: from end-2012 to June 2013 (first MoU update), and then to September 2013 (second MoU update). The condition was finally implemented in September 2013. The project was not sufficiently prepared in advance (the number of targeted tax offices went from 90 in the initial condition to 140 and then finally to 120), which also explains the delays. ‘Parliament adopts legislation to improve the efficiency of the tax administration and controls, implementing recommendations provided by the European Commission and the IMF’ (initial MoU, first programme). This condition was based on the recommendations by the IMF’s technical assistance mission of May 2010 on tax administration. However, the condition was not soundly based on the recommendations: the recommended short-term plan had an implementation timeframe of 12-18 months, while the condition had an unrealistic timeframe of only 3-4 months. Part D : Programmes’ quantitative indicators for tax administration Dec11 Collection of tax debt as of the end of the previous year (million euro) Collection rate for new debt accumulated in the current year New full-scope audits of large taxpayers New temporary audits of large taxpayers Dec13 Dec14 Dec15 946 1099 1518 1561 1641 44 11 % 76 271 19 % 324 590 15 % 411 446 17 % 409 105 65 % 55 % 11 % 13 % 49 % 55 % 29 % 72 % 444 454 693 488 Collection rate in the year for assessed tax revenue from new full audits of major taxpayers Collection rate in the year for assessed tax revenue from new temporary audits of major taxpayers New risk-based audits of self-employed and high-wealth individuals Dec12 404 17 Collection rate in the year for assessed tax revenue from new self-employed and high-wealth individuals’ audits Audit of assets of Tax Collection Offices managers Audit of assets of tax auditors 78 % 22 % 54 72 26 % 104 108 15 % 52 74 Source: Technical Memorandum of Understanding reporting and Greek General Secretariat for Public Revenue activity reports for 2014 and 2015. Target met Target not met Target not set 18 Annex V EVOLUTION OF THE FUNDING GAP METHODOLOGY First programme Second programme Weaknesses Evolution Weaknesses Elements not considered in the initial calculation: In July 2011, the Commission included new elements in the calculation of the government’s financing needs: The funding gap calculation considered only two out of three elements of the Official Sector Involvement (the reduction of the Greek loan facility margin and the commitment by the Eurozone central banks to pass on to Greece the income generated by holding Greek bonds). - external deficit - contingency reserve - risks relating to state guarantees (considered only implicitly through the primary balance forecast). - adjustment of the deficit for cash - cash buffer - settlement of arrears. The decision to return profits on the Securities Market Programme was not reflected in the funding gap calculation. Inconsistencies between the financing conditions (maturities) as presented in the programme document and funding gap calculation. The funding gap calculations made throughout the programme did not provide a breakdown by the contribution of each facility to the overall funding. Evolution New elements considered: - Contribution payment to the ESM capital - Short-term debt reduction (with official funds) - The financing of the PSI cost (cash and accruals) - On the financing side, the funds arising from the additional Official Sector Involvement. The debt of State Owned Enterprises was consolidated in the general government debt. Calculation not based on consolidated general government debt (inappropriate consideration of subentities). Clear breakdown of the cash deficit between the estimated primary cash deficit and interest payments. Reliance on ESA (accrual-based) deficit. Adjustment to cashbased deficits, more appropriate for the funding gap calculation. 19 Annex VI ACCURACY OF THE MACROECONOMIC PROJECTIONS 2010 Actu al% GDP Private consumption Government consumption Investments Exports Imports HICP GDP deflator Unemployment (National Accounts data) 1st EAP 2nd EAP 1st EAP 2nd EAP 1st EAP 2nd EAP 1st EAP 2nd EAP 1st EAP 2nd EAP 1st EAP 2nd EAP 1st EAP 2nd EAP 1st EAP 2nd EAP 1st EAP 2nd EAP 2011 Fore cast % Diff. p.p. -4.0 -1.5 -5.5 -3.8 -19.3 -2.7 -8.2 -7.3 1.5 4.0 -12.0 -2.6 -6.5 -7.0 -20.5 -5.2 -8.0 -7.0 6.1 6.9 1.0 -13.5 1.9 -2.8 -9.1 -0.4 3.5 3.1 1.2 12.0 -0.5 0 0.8 16.7 -23.5 1.2 -6.6 2.8 -6.0 -6.1 -9.4 4.7 2013 Fore cast % Diff. p.p. 1.1 -8.4 -4.7 -2.6 1.0 -9.0 14.7 1.3 2.0 -0.4 23.0 Green figures: favourable variances (forecasts vs actual). Red figures: unfavourable variances (forecasts vs actual). 2014 Fore cast % Diff. p.p. 2.1 -5.3 -5.7 -2.3 -6.0 0 -11 5.0 -2.6 -20.9 -6.2 -17.3 5.7 -4.5 3.2 -2.0 -1.5 -7.6 -5.1 -4.0 1.2 -0.2 -6.5 -9.4 2.2 -1.9 0.0 -3.2 1.1 -3.4 1.5 1.0 -1.4 -0.7 0.3 15.2 7.8 17.9 5.1 -2.5 25.9 Fore cast % Diff. p.p. 2.1 -1.4 2.5 -1.8 1.2 -0.7 0.9 -0.4 0.0 -2.6 -4.7 2.1 1.2 -4.0 10.3 -13.1 6.7 0.8 7.0 0.5 2.1 5.6 2.4 5.3 0.9 -2.3 0.1 -1.5 1.0 -3.2 0.0 -2.2 14.1 10.8 16.7 8.2 0.5 -1.1 -1.2 -1.0 -5.5 -9.5 3.0 1.1 -10.5 6.9 -16.3 7.3 -5.1 5.5 -3.3 1.5 -3.4 0.0 -1.9 0.7 -1.6 -0.9 -0.5 Actu al% 0.7 -2.3 1.0 -0.5 Actu al% -3.2 -8.0 0.0 -10.3 Actu al% -7.3 -4.5 3.4 -3.4 12.0 Diff. -9.7 4.9 0.7 2012 Fore cast % -9.1 -6.5 -4.2 Actu al % -2.6 -2.8 7.5 7.7 -1.4 -0.3 -0.6 0.7 -3.2 -0.4 -2.1 14.8 11.1 17.8 8.1 -2.2 24.9 20 Annex VII DELAYED AND NON-COMPLIED CONDITIONS Labour Condition Final MoU assessment Compliance status Set of conditions on wage bargaining, working time flexibility and employment protection legislation. 1st Programme 3rd review Reforms partially delayed at first assessment (2nd review, 1st Programme). Partial compliance at 3rd review with respectively three or five months delay. Non-complied sub-conditions addressed in subsequent review. Government adopts legislation to remove impediments for greater use of fixed-terms contracts. 1st Programme 5th review Not-observed at deadline (3rd review of the first programme). Following modifications of the condition, assessed as complied with 10 months delay in the 5th review of the 1st programme). Government amends current legislation (Law3846/2010) to allow for a more flexible working-time management […]. 1st Programme 5th review Not-observed at deadline (3rd review of the first programme). Following modifications of the condition, assessed as complied with 10 months delay in the 5th review of the 1st programme). Government amends Law 1876/1990 (Articles 11.2 and 11.3) to eliminate the extension of sector and occupational agreements to parties not represented in negotiations. Government simplifies the procedure for the creation of firm-level trade unions. 2nd Programme Initial MoU Not-observed at deadline (3rd review of the first programme). Following modifications of the condition, assessed as complied with 18 months delay in the second programme. 2nd Programme Initial MoU Assessed as “ongoing” at deadline (4th review of 1st programme). Assessed as complied with a 12-month delay through the establishment of the “union of persons” (2nd programme). The Government will engage with social partners in a reform of the wage-setting system at national level (…) The proposal shall aim at replacing the wage rates set in the national general collective agreement with a statutory minimum wage rate legislated by the government. Business environment 2nd Programme 4th review Assessed as “ongoing / initiated” at deadline (1st review of 2nd programme). Implemented with a 21month delay (April 2014) through several laws. Export facilitation Condition Final MoU assessment Compliance status 21 Government takes measures, in line with EU competition rules: to facilitate FDI and investment in innovation in strategic sectors (green industries, ICT etc...) through a revision of the Investment Law, action to fast-track large FDI projects Government carries out in depth evaluation of all R&D and innovation actions, including in various Operational Programmes, in order to adjust the national strategy. 1st Programme 3rd review Observed in 3rd review of 1st programme, compliance reached with a 4-month delay. 1st Programme 2nd review 2nd Programme 1st review Observed in 2nd review of 1st programme, assessed as complied with a 2-month delay. Not-observed at deadline (3rd review of 1st programme), and also in the 4th review. Assessment in the 5th review: “not applicable yet” Government abolishes the requirement of registration with the exporter’s registry of the chamber of commerce for obtaining a certificate of origin. Government presents a plan for a "Business-Friendly Greece" to tackle 30 remaining restrictions to business activities, investment and innovation. The plan identifies hurdles to innovation and entrepreneurship − ranging from company creation to company liquidation - and presents the corresponding corrective actions. 2nd Programme 1st review 2nd Programme 1st review Assessment in the 2nd programme: “ongoing” “observed” in the 1st review of the 2nd programme with a 2-year delay. Reported as “ongoing“ at deadline (5th review of 1st programme). Assessed as “complied” with a 1-year delay in the 1st review of the 2nd programme. “Partially observed” at deadline (5th review of 1st programme). Assessed as “complied” with a 1-year delay in the 1st review of 2nd programme. Liberalisation of restricted professions and implementation of the Services Directive An audit is launched to assess to what extent the contributions of lawyers and engineers to cover the operating costs of their professional associations are reasonable, proportionate and justified. The Government also adopts legislation to: - reinforce transparency in the functioning of professional bodies by publishing on the webpage of each professional association (…) - the rules regarding incompatibility and any situation 2nd Programme 1st review “Partially observed” at deadline (5th review of 1st programme). Assessed as complied with a 1-year delay in the 1st review of 2nd programme. 2nd Programme 2nd review Legislation “partially observed” at deadline (1st review of 2nd programme). Assessed as “complied” with a 9-month delay in 2nd review of 2nd programme. 22 characterised by a conflict of interests involving the members of the Governing Boards. Government adopts legislation to open up restricted professions including the legal profession, to remove (…) the effective ban on advertising, Government ensures the effective implementation of EU rules on recognition of professional qualifications and compliance with ECJ rulings (...) Under the Services Directive, the government finalizes the review (screening) of existing sectoral legislation, and provides a list of restrictions that are being abolished or amended as a result. Government ensures that the point of single contact distinguishes between procedures applicable to service providers established in Greece and those applicable to crossborder providers (in particular for the regulated professions). The Government also ensures that the electronic point of single contact is operational with a user-friendly internet portal which allows common procedures to be completed by electronic means with the necessary forms available online and recognising electronic signatures in accordance with Decision 2009/767/EC. Government: ensures adequate links between the points of single contact and other relevant authorities (including professional associations); 2nd Programme 3rd review Assessed as complied with a 2-year delay in 3rd review of 2nd programme. 2nd Programme 2nd review Implementation partially observed at deadline (3rd review of 1st programme). Assessed as complied with a delay of 29 months in 2nd review of 2nd programme. 1st Programme 3rd review Flagged as observed at deadline (3rd review of 1st programme) but the report shows that certain areas of the policy are not covered. Delay estimated at more than 6 months. 2nd Programme 1st review Reported as ongoing at deadline (3rd review of 1st programme). Assessed as complied with a 2-year delay in 1st review of 2nd programme. 2nd Programme 3rd review Conditions reported as not-observed at deadline (2nd review of 1st programme), it was then redrafted and divided into various stages. Assessed as complied with a 33-month delay in the 3rd review of 2nd programme. 1st Programme 3rd review Reported as “ongoing” at deadline (3rd review of 1st programme). Assessed as complied in 4th review of 2nd programme as of 5th September 2013 (2 years and 9 months of delay). 23 Government adopts legislation on a limited number of priority service sectors identified in Q4 2010. Government specifies, for a limited number of priority service sectors, a timetable for adopting sectoral legislation by end Q4 2011 that ensure full compliance with the requirements of the Services Directive. The point of single contact is fully operational and the completion of procedures by electronic means is possible in all sectors covered by the Services Directive. 2nd Programme Initial MoU Assessed as “ongoing” in 5th review of 1st programme, following several modifications in the 3rd and 4th reviews. Compliance was achieved in March 2012 with a 9month delay. 2nd Programme 3rd review Not-observed at deadline (5th review of 1st programme). Condition was divided into three parts in 3rd review of 2nd programme, where it was again reported as not-observed. Assessed as fully complied with 2 years and 3 months of delay. Final MoU assessment 1st Programme 5th review Compliance status Financial sector reforms Condition The Bank of Greece (BoG) will require the banks whose capital base has fallen short of regulatory requirements to take appropriate actions via capital injections or restructuring. Should it include the involvement of other local financial institutions, the Bank of Greece will submit a financial impact analysis and legal opinion. The BoG and the Hellenic Financial Stability Fund (HFSF) complete a memorandum of understanding to further strengthen their cooperation, including sharing of appropriate supervisory information. The BoG commits to develop an implementation plan outlining further steps to improve collections and establish targets, in order to ensure an effective utilisation of the enhanced tools. The Government commits to: Adopt definitions for terms as "acceptable living expenses" and "cooperative borrowers", as Reported as “partially observed” due to the length and complexity of procedures. Condition has not been followed up in the 2nd MoU. 2nd Programme Initial MoU Reported as “partially observed” at deadline (5th review of 1st programme). Condition fulfilled with a 9-month delay in the 2nd Memorandum of Understanding. 2nd Programme 4th review Reported as “not observed, pending”. No further evaluation due to the end of the 2nd programme. The condition was open to interpretation. 2nd Programme 4th review Reported as “not observed, pending”. No further evaluation due to the end of the 2nd programme. 24 guidance for the judiciary and banks, with a view to protect vulnerable households. Continue monitoring closely the resolution of distressed debts for households, SMEs, and corporates. The Government commits to build on the significant achievements toward reforming insolvency regimes, by taking the following steps: (i) Established a working group to identify ways to improve the effectiveness of debt resolution processes for households, SMEs, and corporates. (ii) To this end, the Government will, by in consultation with EC/ECB/IMF staff, identify key bottlenecks and (iii) The Government with technical assistance commits to propose concrete steps for the enhancements in this area. The BoG will issue in consultation with banks and EC/ECB/IMF staff, a time-bound framework for banks to facilitate settlement of borrower arrears using standardized protocols, based on the review of banks' distressed credit operations. These (MEFP) include assessment procedures, engagement rules, defined timelines, and termination strategies. Public administration Condition The Government presents an annual better regulation plan (as provided for in Art. 15 of law 4048/2012) with measurable objectives to simplify legislation (including through codification) and to eliminate superfluous regulations. (December 2013) The Authorities will complete shifting at least another 12 500 ordinary employees to the 2nd Programme 4th review Reported as “not observed”. No further evaluation due to the end of the 2nd programme. 2nd Programme 4th review With regard to (ii) and (iii): Reported as “not observed, pending”. No further evaluation due to the end of the 2nd programme. 2nd Programme 4th review Reported as “not observed, pending”. No further evaluation due to the end of the 2nd programme. Final MoU assessment 2nd Programme 4th review Compliance status Not observed, delayed. No further evaluation due to the end of the 2nd programme. 2nd Programme 4th review Not observed, pending. No further evaluation due to the end of the 2nd programme. 25 scheme (...) will have their wages cut to 75 percent. The Authorities will establish quarterly minimum targets for the mobility scheme for 2014. 2 000 more exits of ordinary employees. The Authorities will complete staffing plans for all general Government entities, to be adopted by the governmental council for reform progressively and at the latest by December 2013. The Authorities will complete the assessment of individual employees for the purpose the mobility scheme. The Authorities will reflect the HR strategy in legislation. This legal act will aim at ensuring institutional continuity and higher levels of efficiency in the public administration, and provide a basis for evaluating and developing the competences of the senior management and the staff at large The Authorities take action to consolidate the current preparatory work into a comprehensive and endorsed national e-Government strategy, (…) adopted by the governmental council for reform. Develop an action plan for the assessment of all public entities, including all Extra- Budgetary Funds and SOEs under Chapter A (December 2012). The action plan (…) completed by December 2013. 2nd Programme 4th review Not observed, pending. No further evaluation due to the end of the 2nd programme. 2nd Programme 4th review 2nd Programme 4th review Pending. No further evaluation due to the end of the 2nd programme. Not observed, pending, new deadline for March 2014. No further evaluation due to the end of the 2nd programme. 2nd Programme 4th review Delayed, by 12 months. No further evaluation due to the end of the 2nd programme. 2nd Programme 4th review Not observed. Delayed by seven months (April 2014). No further evaluation due to the end of the 2nd programme. 2nd Programme 4th review The e-Gov Strategy was completed and adopted by the governmental council for reform on 27/03/2014, with 6 month delay from original deadline. 2nd Programme 2nd review Not observed, delayed. No follow up. The Annex is based on a sample of conditions subject to the audit. 26 Annex VIII PUBLIC ADMINISTRATION Part A: Success factors for administrative capacity-building Success factors Process of cultural and organisational change Involvement of stakeholders and civil society ECA’s assessment Not addressed Not addressed Clear methodological and technical approach Not addressed Political commitment Addressed Clear definition of responsibilities Addressed Exchange of best practices at EU level Monitoring and evaluation techniques Not addressed Not addressed Continuity and stability of reforms Not addressed Description A communication plan for public administration reform was included in the MoU conditions after a delay of one year in 2013 and was delivered in 2014. In contrast to other areas of the EAP (i.e. labour market reforms), no exchange of views was organised with stakeholders (local scientific associations, the National School of Public Administration and business associations). The programme lacked a strategic plan for setting conditions on the public administration reform. A strategy and an action plan were required from the Greek government in 2013. They were delivered in 2014. The action plan was not implemented due to a change of government. A new action plan for PAR was included in the MoU conditions in 2015. A high-level political steering committee for public administration reform was included in the MoU conditions. The Ministry of Administrative Reform and E-governance was in charge of the reform process and technical assistance was offered to help steer the reform. MoU conditions did not ensure that PAR measures incorporated good practices. The Commission monitored the achievement of EAP quantitative targets but could not monitor the progress of or evaluate the actual results achieved by the structural reforms from a qualitative point of view. Regarding the reorganisation of the central public administration, a lack of adequate key performance indicators (KPIs) did not facilitate the Commission from verifying the achievement of efficiency objectives such as streamlined administrative procedures and the elimination of overlaps. Structural conditionality measures were not implemented through Administrative Reform Operational Programme projects, but relied heavily on political will. This created a business continuity risk that did materialise. Due to political instability and frequent reshuffles, 10 different ministers led the reorganisation of the public administration within a period of seven years, thus creating a start-and-stop push to reforms. Criteria based on: European Commission, Promoting Good Governance, 2014, p. 6. 27 Part B: OECD’s key recommendations, key reforms; steps for immediate action and the follow up by the programme Key recommendation 1. There is no evident overall strategic vision to provide purpose and direction for the long-term future of Greek society and the economy, or for the short-, medium- and long-term measures to be implemented. Key reforms 2. Pervasive issues of corruption i. can be linked to political and public administrative culture, and its opaque, entangled systems. 3. The Greek Government is not “joined up” and there is very little coordination, thereby compromising reforms that require collective action (i.e. most of them). Establish an HR strategy that is based from top to bottom on non-political appointments and meritocratic criteria, relying on more independent and stabilised structures, building on and clarifying the reforms that have been initiated in this direction. (Included in the programme conditions in 2013 and again in 2015) iv. Establish a strategy to simplify the complex legal and administrative framework and make it more transparent. (Not included) 4. Implementation of policies and i. reforms is a major and debilitating weakness, due to a combination of weak central supervision and a Establish a strategy to address the issues that block the implementation of reforms. (Not included) Steps for immediate action i. Establish a roadmap, milestones and monitoring system to track progress; (Not included in MoU conditions) ii. Identify key players across the administration (central and local) for effective reform; (Not included in MoU conditions) iii. Shape and implement a strategy for regular communication on reform progress, both internally and for the general public. Consider how this needs to be linked to fiscal statements. (Included in MoU conditions in 2013). Establish an information and communication technologies plan to secure interoperability between ministry systems and boost data collection and sharing, starting with core ministries and buildings (pending rationalisation of the latter). (Included in the MoU conditions in 2013. An ICT strategy was delivered in April 2014). Require any new law or policy to include an implementation plan, based on milestones and quantitative fact-based results 28 Key recommendation culture that favours regulatory ii. production over results. ii. iv. Key reforms Monitor reform implementation, using a measurement system that identifies policy priorities and expected results, by establishing indicators/thresholds/best international practice, where feasible. (Not included) Strengthen the structures that link the central administration to the rest of the public sector. (Not included) Identify and enable leaders in the rest of the public sector to deliver on key policy initiatives. (Not included) 6. Human resource management requires equally urgent attention to strengthen the civil service and promote mobility. 7. There are crucial shortcomings in data collection and management, which stand in the way of effective and evidence-based reforms. 8. The Greek administration is hampered by a complex legal framework which discourages initiative, focuses on processes rather than policy, and blocks the progress of reform. Establish a strategy to address data collection and management, with appropriate institutions, funding, and training at all levels of the administration. Implement a government-wide knowledgemanagement system. (Not included) Address the underlying issues that drive the continuous development and use of laws and processes in order to simplify legal structures and processes. Identify and analyse those parts of the legal framework which require reform in order to move the administration’s focus from formal compliance with detailed requirements to the achievement of strategic objectives and policies. (Not included) Source: Greece: Review of the Central Administration, OECD 2011. Steps for immediate action indicators, and to clearly identify those who need to play a part in the implementation process. (Not included) Establish the necessary links between HR and budget reforms. Programme budgeting based on policy objectives to be achieved should be clearly linked to objective-based performance management. The latter should in turn be clearly connected to individual performance appraisals. (Performance budgeting was not applied) Identify essential data for collection by ministries via the strategic central ministry units proposed under issue 3 above. (Not included) 29 Annex IX FINANCIAL SECTOR REFORMS PART A: Programme strategies and key conditions for the financial sector 1st Adjustment Programme Background and rationale: The first programme focused primarily on public finances rather than the financial sector, as was subsequently the case in other Eurozone programme countries. The programme approach was to maintain confidence and avoid spill-over effects from the sovereign debt. As a result, the first programme initially comprised only three financial-sector conditions that were largely inspired by the IMF’s findings for the 2009 Article IV consultation. Key conditions: The main financial-sector condition was to establish the Hellenic Financial Stability Fund (HFSF) with the aim of contributing to the financial stability of the Greek banking system, as the banks were viewed as vulnerable to the economic downturn and to an adverse feedback loop from the sovereign. The fund was set up in July 2010 with 10 billion euros for capital interventions. However, only 1.5 billion euros was actually injected into the Fund during the programme out of which 0.2 billion euro was finally used. The other conditions in the programme were the intensification of supervisory practices by the Bank of Greece and a commitment to review insolvency legislation. 2nd Adjustment Programme Background and rationale: Due to the restructuring of the Greek sovereign debt under the PSI, the losses incurred by all Greek banks (37.7 billion euros in total) and the recession’s impact on asset quality, recapitalisation and resolution were key financial-sector conditions in the second programme and remained significant throughout. Key conditions: Fifty billion euros were earmarked for bank recapitalisation costs in order to offset the expected PSI-related losses, deal with existing and future credit losses, and resolve banks that were no longer deemed commercially viable. The programme envisaged improving supervision and regulation, as well as establishing a stronger governance framework for banks recapitalised by public funds. As the second programme progressed, strengthening private debt restructuring frameworks and banks’ NPL management capacity were identified as higher priorities. 3rd Adjustment Programme Background and rationale: The political crisis that ensued in December 2014 returned Greece's banking system to a situation of severe stress. Liquidity came under immense pressure due to continued significant deposit outflows and the elimination of wholesale funding. Failure to complete the final review of the second programme consequently forced the ECB to maintain the ceiling for emergency liquidity assistance (ELA) to Greek banks at the level decided on 26 June 2015, which in return prompted the Greek authorities to impose a bank holiday, followed by capital controls. Key conditions: The programme supported restoring the stability of the financial system, the aim being to normalise liquidity, recapitalise the systemic banks, enhance governance of both the HFSF and the banks, and address persistently high non-performing loans. To this effect, a buffer of up to 25 billion euros was set aside to address Greek banks’ potential recapitalisation and resolution needs that had to be completed before the end of 2015 due to the impact of the Bank Recovery and Resolution Directive (BRRD), thereby supporting depositor confidence in the system. 30 PART B – The bank recapitalisations under the programmes Bank In March 2012, the implementation of the PSI programme, one of the biggest international recapitalisation debt-restructuring deals affecting about 206 billion euros of Greek government bonds, resulted in a 37.7 billion euros loss for all Greek banks, wiping out their entire capital base. of 2013 As a result, the four larger banks had to be recapitalised by 28.6 billion euros (see Table 1). Table1 - Key figures of all bank recapitalisations under the programmes Bank A Bank B Bank C Bank D Total 2013 (Recap. 1) - capital needs 4,571 8,429 9,756 5,839 28,595 ● HFSF investment 4,021 6,985 8,677 6,700 26,383 ● Private sector investment 0,550 1,444 1,079 0,000 3,073 Total 4,571 8,429 9,756 6,700 29,456 Amounts in € billion unless is stated HFSF shareholding pre-recap n.a. n.a. n.a. n.a. HFSF shareholding post-recap 83,7% 81,1% 84,4% 98,6% ● Baseline scenario 0,262 0,425 2,183 2,945 5,815 ● Adverse scenario 0,560 0,757 2,502 4,980 8,799 ● HFSF investment 0,000 0,000 0,000 0,000 0,000 ● Private sector investment 1,200 1,750 2,500 2,864 8,314 Total 1,200 1,750 2,500 2,864 8,314 HFSF shareholding pre-recap 81,7% 80,9% 84,4% 95,2% HFSF shareholding post-recap 69,9% 67,3% 57,2% 35,4% 2015 (Recap. 3) - capital needs 2,743 4,933 4,602 2,122 14,400 ● HFSF investment 0,000 2,720 2,706 0,000 5,426 ● Private sector investment 1,552 1,340 0,757 1,621 5,270 ● LME / Burden sharing 1,011 0,602 1,019 0,418 3,050 ● Capital actions SSM 0,180 0,271 0,120 0,083 0,654 Total 2,743 4,933 4,602 2,122 14,400 35,41% 2014 (Recap. 2) - capital need HFSF shareholding pre-recap 66,24% 66,93% 57,24% HFSF shareholding post-recap 11,01% 26,42% 26,12% 2,38% HFSF shareholding post-recap (incl. prefs) 11,01% 26,42% 40,39% 2,38% * Potential errors due to roundings ** Includes c. €0.9 billion for a bridge bank *** The quoted figures relate to the initial HFSF investments and do not take into account subsequent reductions Source: Hellenic Financial Stability Fund. With the aim to reduce the high cost of sweeteners to the private sector and avoid possible full nationalisation of Greek banks, the partners proposed a recapitalisation framework aiming to (a) mitigate the disincentives created by prevailing market conditions and future uncertainties, and (b) minimise losses to the taxpayer by providing only upside leverage for the private sector that would simultaneously allow the HFSF to exit without bearing additional losses. This was considered a critical part of the ongoing strategy for reorganising the banking system and preserving the banks’ business autonomy. It was agreed to set a minimum threshold for the private-sector contribution at 10 % of an individual bank's capital needs. Furthermore, it was agreed that if the private sector contributed at least the minimum 10 % of capital needs, the HFSF would cover the remaining 90 %, but would only receive 'B Shares' (i.e. with suspended voting rights except for certain key veto powers) in exchange, thereby preserving the private management of the bank. If the private sector was not willing to cover the minimum 10 % of capital needs, the HFSF would recapitalise the bank fully and receive full voting rights, thereby diluting the existing shareholders and in effect nationalising the bank. As banks were insolvent, free warrants were also offered as a sweetener to encourage private investors to become involved in the recapitalisation. The warrants were deemed complex, while providing significant sweeteners to private sector (as of June 2013, the market value of the free warrants offered to private investors was estimated at 1.7 billion 31 euros). During May and June 2013, the four largest Greek banks completed their share capital increase. Three out of the four banks managed to raise more than the minimum 10 % of their required capital needs, i.e. almost 3.1 billion euros from private sources. The HFSF thus ended up contributing the remaining amount of 25.5 billion euros to these banks (not including 0.9 billion euros for a bridge bank), and provided warrants to private investors for its shares, as per the agreed recapitalisation framework. Banks decided to raise more capital by means of ordinary shares, thereby forgoing the option of issuing expensive and potentially dilutive contingent convertible bonds (CoCos). Following the failed merger with another bank and in the light of increased deposit outflows due to speculative spill-over effects stemming at that time from the Cypriot bail-in decision, the fourth bank opted for an immediate and full recapitalisation via the HFSF rather than attempting to raise the minimum 10 % of capital needs from private sources. Given the time remaining until the end of April (second programme’s deadline) and the very high valuation of the shares to be issued, interest from private investors would be limited. As a result, in May 2013 the HFSF injected 5.8 billion euros into the bank, becoming the main shareholder (98.6 %) with full voting rights, unlike the other three systemic banks in which the HFSF had received shares with restricted voting rights. Bank The protracted recession had a negative impact on Greek banks’ liquidity, balance sheets and recapitalisation financial results. Following a new capital needs assessment conducted by the Bank of Greece, in compliance with a condition under the second programme, the second of 2014 recapitalisation of the four largest Greek banks was finalised until June 2014. In particular, 8.3 billion euros were raised entirely through the private sector (see Table 1). Following the second recapitalisation, two of the banks were able to raise also funds through bond issuance. In the case of a bank, the share capital increase of 2.9 billion euros for the baseline scenario took place in accordance with revised Law 3864/2010, according to which the HFSF could act only as a backstop. The share capital increase involved a prior commitment by a cornerstone investor to a specific size and a specific price. In mid-April 2014, the HFSF received one binding offer at 0.30 euros per share for a total of 1.3 billion euros from a consortium of investors. The remaining amount was covered during the book-building process. The final offer price was set at 0.31 euros per share (at a discount of 23 % over the last closing price; see footnote 54), with the cornerstone investor raising its offer to match that price. The share capital increase was completed by the end of April and the amount was covered in full by institutional investors and private investors through a private placement and a public offering. As a result, the HFSF’s shareholding was reduced from 95.2 % to 35.4 %. Bank In 2015 the protracted negotiations between the Greek authorities and the institutions, recapitalisation along with the substantial outflow of deposits and the constantly increasing NPLs against the of 2015 backdrop of a worsening economic climate and capital controls necessitated a third bank recapitalisation. The capital needs assessment conducted by ECB found the four largest Greek banks to be short of 14.4 billion euros under the adverse scenario (see Table 1). Subsequently, the banks submitted their respective capital plans to the ECB, detailing how they intended to address their capital shortfalls. Given the context of the Bank Recovery and Resolution Directive (BRRD), the state aid framework, and the Eurogroup’s guidance of 14 August 2015, the priority of key objectives for the 2015 recapitalisation was to address any capital shortfalls, avoid a depositor bail-in by completing recapitalisation in 2015, avoid nationalisation of the Greek banks, minimise state 32 aid by maximising private investment, and avoid undue dilution of the HFSF’s shareholdings. The two banks, which had the lowest capital needs, managed to cover the entire amount of the adverse scenario from private investors (3.2 billion euros) and by voluntarily converting all subordinated and senior bondholders into equity. The other two banks, which had higher capital needs, raised the amount of capital required only under the asset quality review (AQR) and baseline scenario from private investors (2.1 billion euros) and by converting subordinated and senior bondholders into equity. It should be noted that the European Bank for Reconstruction and Development (EBRD) and the World Bank’s International Finance Corporation (IFC) also invested in the four banks. The remaining amount of 5.4 billion euros was provided via HFSF from programme funds HFSF (i.e. 1.3 billion euros in new shares and 4.1 billion euros as CoCos). 33 Annex X FINANCIAL STABILITY The EBA’s risk indicators for the Greek and EU banking systems Weighted Averages 2013 2014 2015 2016 Greece Solvency Total capital ratio Tier 1 capital ratio CET1 ratio CET1 ratio (fully loaded) Greece EU Greek banks' ranking in the EU-28 12.7 % 12.4 % … … 14.1 % 13.9 % 13.9 % 5.6 % 16.4 % 16.3 % 16.3 % 15.0 % 17.1 % 17.0 % 17.0 % 15.9 % 18.5 % 15.5 % 14.2 % 13.6 % 22 14 12 13 Credit Risk and Asset Quality Ratio of non-performing loans and advances (NPL ratio) Coverage ratio of NPLs and advances Forbearance ratio for loans … 39.7 % 46.2 % 45.9 % 5.1 % 28 … … 43.8 % 14.2 % 48.5 % 19.8 % 48.2 % 23.2 % 44.6 % 3.2 % 12 27 Ratio of non-performing exposures (NPE ratio) … 33.9 % 37.3 % 38.5 % 4.4 % 27 … … 62.2 % (11.1 %) (1.0 %) 60.9 % (25.5 %) (2.5 %) 59.8 % (7.7 %) (0.9 %) 51.9 % 3.3 % 0.2 % 65.7 % 26 28 14 78.3 % 84.1 % 86.4 % 82.0 % 57.9 % 28 14.8 % 14.5 % 10.9 % 12.4 % 27.2 % 28 … (4.1 %) 2.1 % (2.9 %) 6.0 % 26 Profitability Return on equity Return on assets Cost-to-income ratio Net interest income to total operating income Net fee income and commission income to total operating income Net trading income to total operating income Net interest income to interest bearing assets Balance Sheet Structure and Liquidity Loan-to-deposit ratio … 2.8 % 2.5 % 2.9 % 1.5 % 7 100.7 % 109.2 % 129.7 % 120.2 % 118.4 % 21 Leverage ratio (fully phased-in) Leverage ratio (transitional) Debt-to-equity ratio … … 1132.0 % … … 999.0 % … … 837.8 % 10.0 % 10.7 % 754.0 % 5.2 % 5.5 % 1440.7 % 22 27 7 Asset encumbrance ratio … 26.0 % 47.1 % 43.3 % 26.3 % 25 Liquidity coverage ratio … … … 0% 141.1 % 28 34 NB. EBA’s NPL ratio is based on the broader definition for NPEs and refers to loans and advances; EBA’s NPE ratio also includes debt securities (mainly bonds). Source: European Banking Authority, Risk Dashboard Data (from a sample of EU banks; 198 in 2016). 35 Annex XI SUSTAINABLE GROWTH GDP GROWTH (%) 2009 2010 2011 2012 2013 2014 2015 2016 -4.3 -5.5 -9.1 -7.3 -3.2 0.4 -0.2 0 Target (MoU forecast, May 2010) -2.0 -4.0 -2.6 1.1 2.1 2.1 Target (MoU forecast, March 2012) -4.7 0.0 2.5 Target (MoU forecast, April 2014) -0.1 3.3 INFLATION – HICP (%) 2009 2010 2011 2012 2013 2014 2015 1.3 4.7 3.1 1.0 -0.9 -1.4 -1.1 2016 Target (MoU forecast, May 2010) 1.9 -0.4 1.2 0.7 0.9 Target (MoU forecast, March 2012) -0.5 -0.3 0.1 Target (MoU forecast, April 2014) -0.8 0.3 UNEMPLOYMENT RATE (%) 2009 2010 2011 2012 2013 2014 2015 9.6 12.70 17.9 24.5 27.5 26.5 24.9 Target (MoU forecast, May 2010) 12.0 14.7 15.2 14.8 2016 36 Target (MoU forecast, March 2012) 17.9 17.8 16.7 Target (MoU forecast, April 2014) 24.5 22.5 YOUTH UNEMPLOYMENT (%) 2009 2010 2011 2012 2013 2014 2015 25.7 33.0 44,7 55.3 58.3 52.4 49.8 2016 Target (ECA based on MoU forecast, May 2010) 1: 31.2 38.2 39.5 38.5 LONG-TERM UNEMPLOYMENT (%) 2009 2010 2011 2012 2013 2014 2015 3.9 5.7 8.8 14.5 18.5 19.5 18.2 2016 Target (ECA based on MoU forecast, May 2010) 2: 5.4 6.6 6.8 6.6 EMPLOYMENT RATE (%) 2009 2010 2011 2012 2013 2014 2015 65.6 63.8 59.6 55.0 52.9 53.3 54.9 2016 Target (Europe 2020) 70.0 WOMEN EMPLOYMENT (%) 2009 2010 2011 2012 2013 2014 2015 52.9 51.8 48.7 45.2 43.3 44.3 46.0 2016 1 Adjusted MoU employment forecast, assuming stable ratio of youth/overall unemployment at 2010 level. 2 Adjusted MoU employment forecast, assuming stable ratio of long-term/overall unemployment at 2010 level. 37 Target (ECA; based on Europe 2020) 3: 59.3 EXPORT GROWTH 2009 2010 2011 2012 2013 2014 2015 -20.2 10.8 5.8 3.7 0.6 5.2 -8.7 2016 Target (MoU forecast, May 2010) 3.4 7.1 6.7 9.3 Target (MoU forecast, March 2012) 2.9 5.2 6.9 Target (MoU forecast, April 2014) 0.4 3.4 5.5 5.0 2016 EXPORT SHARES (5-YEAR CHANGE, %) 2009 2010 2011 2012 2013 2014 2015 -10.2 -14.0 -15.5 -24.7 -25.2 -18.0 -20.6 -6.0 -6.0 Target (based on MIP threshold) -6.0 -6.0 -6.0 -6.0 -6.0 FDI (% OF GDP – AVERAGE OF INFLOW AND OUTFLOW) 2009 2010 2011 2012 2013 2014 2015 0.8 0.2 0.4 0.7 1.2 1.1 0.6 2.1 2.1 2016 Target (ECA) EU-28 average in 2010 2.1 3 2.1 2.1 2.1 The overall Europe 2020 employment target (70 %), adjusted for the employment gender differential at EU average level in 2015 (i.e. 0.85, given female employment at 64.3 and male at 74.9). REPLIES OF THE COMMISSION TO THE SPECIAL REPORT OF THE EUROPEAN COURT OF AUDITORS "THE COMMISSION'S INTERVENTION IN THE GREEK FINANCIAL CRISIS" EXECUTIVE SUMMARY I. The Commission welcomes this European Court of Auditor’s (ECA) performance audit of the Commission’s involvement in the Greek financial assistance programmes financed by the different mechanisms put in place over time since 2010. The Commission is open to constructive criticism and well-founded recommendations on how improvements can be made in the design and implementation of financial assistance programmes. The Commission intends to build on the changes identified and, in some cases, already under implementation in order to bring about further improvements, as set out in its response to the audit recommendations. Far-reaching changes have been taken up under the ESM framework and implemented already within the current ESM stability support programme for Greece. In the context of the ESM stability programme for Greece, the Commission has put a yet greater emphasis on social cohesion, growth and employment. Also, for the very first time, the Commission has conducted a social impact assessment (see also reply to paragraph IV. of the Executive Summary below). The Commission would like to recall the significance of the political and economic changes that impacted on decision-making, design and implementation during the programme years. IV. The underlying objectives of the financial assistance programmes is to restore market access and by construction they are established in moments of acute crisis; this is especially true in the case of Greece. There is thus a need to distinguish between the immediate and medium-term objectives of the programmes. Policy conditionality should take into account the capacity of the national authorities to adopt and implement policies that are economically and socially very challenging. The policies adopted in the context of a financial assistance programme provide the framework for sustainable growth and jobs over the long-term, even beyond the time horizon of a programme itself; and there is thus a need to establish comprehensive growth strategies to this end. Furthermore, it is of utmost importance to systematically take into account the economic and political context of each financial assistance programme, in which relevant policy choices were made. For example, the first programme was established against the background of the sovereign having been abruptly cut off from market access. The Commission had to act in a context of extreme and unprecedented uncertainty, a severe liquidity crisis that challenged the stability of the whole financial system, and in the absence of adequate financial assistance instruments at the level of and available to the Euro area. It is moreover crucial to preserve the chronology of events. It is also essential to recognise fully both the different frameworks and the circumstances of the establishment of the ESM stability support programme for Greece. Therein, the Commission acted within a new legal framework (Regulation 472/2013) which dramatically improved the transparency of its work and its democratic accountability, through a reinforced dialogue with the European Parliament and the national Parliaments. The consistency of the programme with the Union goals and policies was also reinforced: by ensuring that the MoU will be consistent with the programme approved by the Council; through explicit references to the Charter of Fundamental Rights as well as other social rights; the taking into account of the national practice and institutions for wage formation as well as the national reform programme of the Member State concerned in the context of the Union’s strategy for growth and jobs; and the requirement that budgetary consolidation efforts in the programme take into account the need to ensure sufficient means for fundamental policies such as education and health care. Consequently, the ESM programme has put a greater emphasis on social cohesion, growth and employment (including active labour market policies); and the Commission, 1 for the first time, conducted a social impact assessment. The new legal framework also created a basis for the Commission to provide technical assistance to Greece to improve its administrative capacity and address problems implementing the programme. V. Whilst the Commission is the addressee of this performance audit for legal/institutional reasons, both the Eurogroup and other Union institutions play a key role that has changed significantly over successive programmes. A good understanding and reflection of the actual governance framework prevailing in each programme or during the design of the current ESM stability support programme, and at the time specific decisions were made is necessary for actions and decisions to be appropriately attributed to the responsible parties at all times. Commission actions were developed and implemented under complex institutional settings that have evolved significantly since 2010 in line with changes in the legal basis. Moreover, the role of the Commission in the negotiation process implies interaction with not only national authorities and other EU institutions, but also multiple international organisations and the Eurogroup. In this respect, the duties conferred on the Commission and the European Central Bank (ECB) within the framework of the European Financial Stability Facility (EFSF) or, subsequently, the Treaty establishing the European Stability Mechanism (ESM), important as they are, do not entail any power to make decisions on their own. Further, responsibility for not only policy choice but also implementation lies with the national authorities. All Memoranda of Understanding have been based on shared conditionality agreed by all institutions, including the International Monetary Fund (IMF), and in agreement with the Greek authorities, after a process of internal discussions and in-depth dialogue. The need to have shared conditionality was pursued by the Eurogroup. This required reaching compromise positions with the other institutions, namely the IMF and ECB. Policy conditionality was designed after detailed internal discussions among the institutions that involved many oral and written iterations. The provision of technical assistance – where relevant – was designed to provide the best possible support to the national authorities to implement the conditionality by the agreed deadlines. One strength of this multi-institutional framework is that different institutions can pool expertise, which often enhanced the quality of the policy design. Moreover, it is the Member States – or ESM Members under the ESM framework – that establish the financing envelope and also decide upon measures related to debt. The ECB, and subsequently the Single Supervisory Mechanism (SSM), have specific responsibilities to ensure financial stability and, as supervisors, take decisions independently and in some cases without sharing confidential, market-sensitive information with other institutions. VIII. The Euro area lacked financial instruments and a legal framework to provide financial assistance to Euro area countries. The first programme, financed by the Greek Loan Facility (GLF), was rapidly established in order to avert the default of the sovereign. In the absence of a framework at the time, the Commission and the Eurogroup relied upon the framework and methodologies developed by the IMF, which was then the international organisation with the mandate and the experience to undertake such programmes. The Commission formally codified its own procedures in 2011. Policy actions were duly prioritised, notably through the joint programme with the IMF. The Commission, inter alia, used the IMF's well known system of 'prior actions' and 'structural benchmarks', which are critical reforms needed to close a review and release a disbursement. These were gradually refined with some additional prior actions in the area of structural reforms, and through the use of milestones. The ESM stability support programme currently underway also introduced the concept of 'key deliverables'. IX. Each institution acts within the legal framework applicable to each programme. XI. The design and implementation of crucial reforms took place in the wider context of the prevailing difficult economic situation as well as severe instability in the financial markets. The successful recapitalisation, substantial restructuring and regulatory and governance reforms 2 undertaken in direct response to an acute crisis and to contain unfolding negative effects allowed for the stabilisation of the entire system. This ensured the achievement of the key objective of the programmes: avert a sovereign default and ensure financial stability. The counterfactual scenario (a financial system collapse) would have brought about far more significant financial, economic and social costs. XII. The absence of political stability created challenges of ownership of the reform agenda over time; this constitutes one of the key elements to be kept in mind when assessing policy outcomes in this area. XIII. Greece experienced recurrent protracted periods of political instability that reignited uncertainties regarding the policy course, commitment to reforms and their effective implementation. However, Greece tapped the markets in April and July 2014, following a period of a steady reform, successful conclusions of reviews, and improved growth prospects. This clearly demonstrates how effective reform implementation is conducive to increased confidence among market participants and a successful return to the markets. First bullet point: A successful return to growth depends on successful programme implementation. Growth could not be achieved without addressing the underlying systemic and structural weaknesses of the Greek state and economy. Given that both the first and the second programmes were interrupted, this ultimate goal was not entirely achieved. Third bullet point: The implementation of the programmes prevented the collapse of the financial system that would have resulted in far more dramatic consequences for the Greek state and for financial stability. However, adverse macroeconomic and political developments, and the protracted implementation of financial market reforms under the programmes contributed to deterioration of the banks' balance sheets. XIV. Please see the Commission replies to the Conclusions and recommendations parts below. INTRODUCTION 6. At the end of June 2015, the Greek government decided unilaterally not to complete the 2nd EFSF programme and called a referendum. 8. See Commission reply to paragraph V. of the Executive Summary (institutional setting). OBSERVATIONS PART I – MANAGEMENT OF THE ECONOMIC ADJUSTMENT PROGRAMMES FOR GREECE 24. Please see the Commission's reply to paragraph VIII. 25. In the context of the first programme, the Commission and the Eurogroup relied upon the framework and methodologies of the IMF – the international organisation with the mandate and the experience to undertake such programmes. The programme design was preceded by a considerable amount of research, covering the elements to be tackled by the reforms as well as introducing criteria for reform prioritisation. Moreover, the Commission was able to draw upon methodologies and analytical tools developed and applied under normal surveillance regimes in areas such as forecast, analysis of the fiscal stance and sustainability, pensions and the benchmarking of structural policies/reforms. 27. The Commission produced analytical documents supporting the policy briefs and corroborating the design of conditions, including in the area of regulated professions. The first few conditions on regulated professions and the services directive, for example, were completed by the Greek government on time while reduced willingness on the Greek authorities' side ultimately induced delays. 3 28. First bullet point: While some conditions were initially expressed in more general terms, setting out a clear end-objective, regular iterations with the authorities then produced more detailed guidance that laid out the means and steps by which such reforms would be implemented. This approach did not lead to proper implementation in some cases. Interpretational issues, however, were not necessarily the main cause for delayed implementation. Second bullet point: Before the Greek authorities commit to any conditions and sign the Memorandum of Understanding, every condition is always discussed and agreed with the authorities, both at technical and political level to ensure a thorough and common understanding and ownership of the reform process to be undertaken. The high level of detail was motivated by a need to identify clearly how the reforms were to be undertaken and address implementation challenges. Moreover, the authorities' administrative capacity was challenged by the number of measures necessary to truly improve efficiency of the tax administration rather than by their level of detail. Limited ownership of the reforms during the second programme was often an issue, notably regarding the staff of the tax administration. 29. See Commission reply to paragraph IV. of the Executive Summary on the objectives of the programme and the general context. The comprehensive set of reforms under the ESM programme has been, moreover, flanked by a set of extraordinary measures to help Greece make best use of available EU funds and technical support under the Commission's initiative set out in the Communication of 15 July 2015 on "A New Start for Jobs and Growth in Greece". 30. The first programme had more elements of horizontal policies, while the number of conditions targeting specific sectors increased in the second programme with conditionality on screening of business restricting regulations and targeting sectors such as building materials and manufacturing in addition to tourism and retail. This conditionality was carried out with the assistance of the OECD competition toolkit. Box 2 – Product market reforms versus labour and tax reforms The Commission was fully aware of the effect of higher indirect taxes on price developments. However, there has been an underlying price adjustment process ongoing in the Greek economy, which has contributed to regaining price competitiveness. The primary objective of the first programme was to restore Greece’s access to market financing, which necessitated a clear focus on fiscal consolidation. Nonetheless, the overall design of the Greek financial assistance programmes was underpinned by an explicit strategy. Structural reforms formed a key part of the programme and were aimed precisely at producing positive impact beyond the programme horizon. 39. In order to benchmark Greece's performance and evolution over time, the Commission relied on indicators that were available from international sources, but were not always fully comprehensive and were only available with a time-lag. This includes OECD product market indicators to measure progress made in reforming regulated professions as well as the World Bank's Doing Business indicators. Moreover, the MoUs included specific conditionality to assess the impact of the reform of regulated professions – including a survey of the 20 largest professions examining the degree of liberalisation, results with respect to new entrants, price changes, etc. 40. During a crisis, it is imperative to prioritise actions, particularly when there is limited institutional capacity to address all actual and potential issues immediately. 4 In the Greek financial assistance programmes, priority was given to recapitalisations that were urgently requested by the supervisors for financial stability purposes. The issue of NPL resolution was addressed as a critical next step, given the increase in NPLs as a result of a protracted crisis period. The process of addressing NPLs is longer and more complex, and requires reforms on several fronts (legal, supervisory, governance, regulatory, etc.) and in different contexts, each presenting its own peculiarities (household debt, SME and bigger corporate business debt). It is recalled that the Asset Quality Review in 2013 included a Troubled Asset Review, which measured banks’ preparedness and capacity to manage NPLs. As regards the HFSF, the Commission has only an observer role. 45. There is no single macroeconomic model that can be readily used for economic forecasts. Moreover, it is crucial to systematically take into account the fast-changing economic and political context in Greece when assessing the timeliness, quality and impact of policies adopted under the programmes. Box 4 – Coordination of projections Throughout the programmes, the mutual impact of both macroeconomic and fiscal projections was estimated through an iterative process in which the impact of new fiscal measures was included in the macroeconomic baseline scenario and the latter was used to estimate the fiscal projections. 48. In view of the high pressure and very short deadlines, minor errors or omissions may have occurred on occasion. Throughout the programmes there was strong and systematic interinstitutional quality and peer review of data and calculations. 50. The amount of EUR 8 billion was considered a roughly acceptable target, taking into account the financing needs after the end of the programme which in turn depends on possible debt measures to be implemented. Moreover, the complexity of the institutional and legal framework, with its multiplicity of actors and decision makers, should be kept in mind. 51. Given the lack of consensus in the economic literature, projections must rely on a few studies that can give guidance to the scale of the impact. Furthermore, macro-fiscal projections have been integrated and this concerned not only the Commission's forecast and not only for Greece. PART II - DESIGN AND IMPLEMENTATION OF THE REFORMS 56. The Commission stresses that the responsibility for not only policy choices but also implementation lies with the national authorities in every financial assistance programme. 57. See Commission reply to paragraph IV of the Executive Summary (objectives and context). 59. See Commission reply to paragraph IV of the Executive Summary (programme objectives). 61. See Commission reply to paragraph IV of the Executive Summary (programme objectives). 63. See Commission reply to paragraph IV of the Executive Summary (programme objectives). Tackling tax evasion and improving tax compliance is clearly an important avenue for reforms in Greece. Nevertheless, it is not one that can be solved quickly. The ESM programme has laid great emphasis on such measures, building on the experience of the prior programmes. Box 6 – Scope of measures A split-payment mechanism for VAT transactions with public institutions has been contested by experts. The split-payment mechanism for VAT transactions with the public sector has been used in Italy, but has been contested by both business and tax practitioners. 5 A more intensive use of electronic fiscal devices requires a tax administration equipped with advanced IT systems, which was not the case in Greece. Box 6 – Initial detail The first programme did not lay down all details of the committed reforms. However, details were further elaborated in the course of the programmes and based on existent experience, in order to facilitate a clear understanding of the authorities and ex-post evaluations. 66. The growing tax debt has been affected by the fact that the write off process has only been put in place in 2014 and is moreover very lengthy (tax debt is first quarantined, if the debtor does not develop sufficient capacity to repay the debt, it is subsequently written off but only 10 years later if no improvement of the solvability of the debtor takes place). It is a considerable technical challenge to devise indicators for monitoring the estimated level of undeclared taxes which can be accurate within the programme review periods. 68. The programmes provided adequate strategic framework for public administration reforms. The immediate priority in this sector was fiscal consolidation, and this explains the sequencing of reforms, from fiscal to structural. The Commission, notably via the Task force for Greece (TFGR) provided exchanges of best practices through technical assistance. The Commission regrets that attempts to involve stakeholders were unsuccessful. Lack of continuity and stability were indeed relevant issues, due to frequent changes in government. (see also reply to paragraph 76). Finally, there was considerable inertia and resistance to change within the Greek administration. 73. The Commission set ambitious deadlines; but the specific example for over-ambitious deadlines is not appropriate. The design and implementation of reforms on the appraisal system under the new programme were not executed because the general implementation of the second programme was effectively halted from Autumn 2014 to August 2015, when the ESM programme was launched. Therefore, it is not possible to say whether the deadlines set under the second programme were too ambitious or not: implementation was not even attempted as the programme was ended before implementation could be completed. Further, the new appraisal system is being implemented under the ESM programme. Time between agreement of conditionality and completion of secondary legislation was ten months (August 2015 to June 2016), one month less than what was provided for in the second programme; while an actual assessment is being done in 2017 because it is based on year-end 2016 data, as it should be in order to get a full picture over time. 78. Key determinants of credit supply were addressed in the programmes, which included policies that addressed bank capital and liquidity, as well as tackling NPLs. The factors behind credit contraction are complex and reflect structural problems in credit markets. The high interest rate charged for new loans is, therefore, largely rather a symptom of a deep recession, due to borrowers' inability to repay their loans. At the same time, the Commission has worked to increase financing from EU and International Financial Institutions (EBRD, EIB, EIF) which includes financing instruments for SMEs. 79. The analysis was a shared view of all institutions involved and was based on information available at that time. It was also based on input from independent experts. The recapitalisation framework was prepared by professional investment bankers, which were consultants to the HFSF, the Bank of Greece and the Ministry of Finance. 6 80. The capital needs in 2013, 2014 and 2015 were determined under the sole responsibility of the supervisors. The Commission and the other institutions merely provided technical support to develop the framework of the stress test. Furthermore, political events and ensuing uncertainty and instability had a significant negative impact on GDP growth, and constitutes thus an element that could not have been anticipated. 81. It is the responsibility of the national authorities to implement a predefined task within the agreed perimeter. The Commission does not interfere in the relationship between the supervisor and the banks. 82. According to EU State aid rules, aid should be limited to the minimum necessary and hence used to cover the capital needs of the bank only if no private market funds became available. As explained in the Commission's State aid decisions adopted in 2014 and 2015, the participation of the HFSF cannot be considered as a market transaction as the Fund did not invest under the same terms and in the same transactions as the other investors. In the 2014 recapitalisation, a mechanism was introduced to calculate a price at which the HFSF was authorised to participate, precisely to avoid excessive dilution of the HFSF. In the 2015 recapitalisation, no minimum price for HFSF participation was determined ex ante due to specific market conditions and, as of a decision of the authorities, the subscription price for the HFSF was determined by a book building process performed and monitored by international experts. These mechanisms helped reduce the losses for the HFSF. Share price developments in the following recapitalisation reflect the extent of the risk undertaken by the private investors amid conditions of heightened uncertainty. 83. In the specific case of Greece, it quickly became obvious to all institutions involved in the programme, that the creation of an AMC would not constitute an adequate solution for a variety of reasons: heterogeneity of NPLs spreading over most economic sectors, extreme cost of such structure, governance concerns, importance of the bank-client relation in a context of widespread strategic default, impact on banks' balance sheets, etc. 84. See Commission reply to paragraph IV. of the Executive Summary (programme objectives). 85. Improving insolvency legislation and judicial capacity is a complex reform that can only be achieved by a multitude of actions. Several legislative changes, including to the Insolvency Code and the creation of an Out Of Court framework, were introduced between 2010 and 2014. Implementation of reforms remains the responsibility of the Greek authorities. This implementation effort was lacking. Moratoria on foreclosures were introduced as a unilateral action by the Greek authorities and prolonged (the last extension was until 30 October 2015). Box 10 It is recalled that ultimately the responsibility for the successful implementation of reforms has been and remains the responsibility of the Greek authorities. Please see reply to paragraph 85. 86. Whereas the market for NPL servicing and sales was established under the ESM programme, the 2nd programme already contained a request to develop a Comprehensive NPL Strategy as a key condition under the MoU; the Greek authorities were, therefore, already required to remove key impediments to the creation of an active NPL market well before the law of December 2015 was adopted. Additionally, the Commission along with other institutions and international partners asked to remove all impediments to the free NPL market (overly bureaucratic procedures, supervisory requests, levies on NPL sales etc.) that contributed to amending the law in the following months in view of the closing of the first review. Finally, the main remaining impediments were removed thereafter, in 2017. 7 87. The ultimate responsibility for the successful implementation of reforms remains with the Greek authorities. Programme conditionality eventually requested the review of the liquidation arrangements and a correction of inefficiencies. The programme targeted an improvement in the way liquidations were organised, and how these would be handled when problems materialised. 88. The focus of the first programme was mainly on the immediate critical fiscal policies, while the banking sector was considered relatively sound at the time. While it is correct that the first programme did not include any direct conditions on bank governance, this issue was indirectly touched upon as part of the programme conditions for the establishment of the HFSF. The Hellenic Financial Stability Fund was already established under the first programme with the objective to strengthen the health and resilience of the banking sector, where governance was one of the major concerns. Bank governance issues gained prominence in expectation of the additional use of public funds to recapitalise banks under the current ESM programme. 89. All the banks have not been nationalised. To unilaterally replace the management of private banks raises legal difficulties in justifying direct intervention within property rights. It is the role of the shareholders to agree on the bank management, as recognised by EU law. Initially, the HFSF was not involved on the grounds that profound governance changes in the midst of a severe crisis would have aggravated deposit outflows, depleted these banks of experienced management and hence created serious financial stability risks. This was the time when many banks were to carry out a capital increase to cover their significant capital needs. In the absence of early progress to enhance the governance of banks' boards, the issue has been addressed under a condition in the MoU of the ESM programme. 90. The additional requirements help ensure that boards of higher quality and free from outside interference are established. Given that Greece faced an unprecedented problem of liquidity, capital and asset quality, it was crucial that the banks' boards hire the best possible experts to address the specific challenges of the Greek banks. 91. The ultimate responsibility for the successful implementation of reforms remains with the Greek authorities. The Commission with the other institutions carefully monitored the less significant institutions; however, ensuring financial stability while focusing on the systemic banks was the priority. Under the ESM programme, less significant institutions were not eligible for public/Programme funds for recapitalisation. Hence, there is no justification by the programme to intervene in these banks as long as they were not in breach of prudential requirements. The institutions had no role in the mentioned audit but welcomed the decision of the supervisors to perform it. Supervisory issues related to the Less Significant Institutions (LSIs) were captured in the Comprehensive Financial Sector Framework, which included the restructuring of the entire banking sector where several LSIs have been resolved or liquidated. 92. The HFSF Law contains detailed rules regarding decisions which are within the remit of the Executive Board and those of the General Council (Art 4). The responsibility for any incorrect implementation solely lies with the HFSF since the institutions have no responsibility for the management of its tasks. 8 93. While not all changes can be attributed to programme requirements, a number of changes were indeed necessary to improve the functioning of the HFSF and were related, in particular, to the need to ensure the independence of the HFSF from outside political influence. While there were potential downsides to such changes, benefits were deemed to outweigh the costs. 94. This situation should be assessed in the wider context of the prevailing difficult situation in the financial markets, which does not allow for swift recovery of investment. The successful recapitalisation, substantial restructuring and regulatory and governance reforms, undertaken in direct response to contain the unfolding negative events, allowed for stabilisation of the entire system. This ensured the achievement of the key objective of the programmes, namely financial stability. The alternative scenario (i.e. financial system collapse) would have brought about significant financial, economic and social costs. 97. While there was clear awareness of the risks stemming from the grey zone, their meaningful assessment would have been impossible given their illegal nature (illegal under-reporting of earnings). The Commission recalls that the negotiations were conducted jointly with other institutions. See Commission reply to paragraph V. of the Executive Summary (institutional setting). 98. See Commission reply to paragraph V of the Executive Summary (institutional setting). 99. Measures in the field of collective bargaining were agreed and undertaken in a gradual and progressive manner. This, however, did not stem from a failure to recognise that the Greek economy is characterised by the prevalence of micro and small enterprises. Rather, it reflects a prudent approach, with subsequent additional interventions – such as the introduction of "unions of persons" allowing firm-level bargaining to take place also in companies without trade union representation, which affected more directly the existing national practice – being taken at a later stage in response to an ever-deteriorating situation. 100. On the topic of labour market reforms, the same themes were included under the ESM programme as in the previous programmes for two main reasons: firstly, some of the measures agreed under the second programme had not been implemented; secondly, the ESM programme foresaw a comprehensive review of the measures that had been implemented in the past – some of which of a temporary nature – to establish whether these were still adequate and necessary. While recognising important delays in implementation, the Commission considers that the timing and sequencing were appropriate. Box 11 The minimum wage set in the national collective labour agreement was reduced by law, and a new framework introducing a statutory minimum wage was established. This reform deeply changed the nature of minimum wage setting in Greece, from collectively agreed (autonomously by social partners) to statutory (set by government). At the same time, the newly established framework was meant to be applied after the programme (i.e. under stable economic and labour market conditions), with the level of the minimum wage being kept frozen over the programme horizon. 102. Third indent: The reference to a quantitative level (15% reduction) was included as an indication of the order of magnitude of the reduction expected in unit labour costs. Developments in the latter variable are the result of complex interactions, which can be influenced but not directly "controlled" by the public authorities. It would therefore be wrong to take this level as a clear target against which to measure success/performance. PART III - ACHIEVEMENT OF PROGRAMME OBJECTIVES 106. When considering whether the financial assistance programmes met their main objectives, it is the responsibility of the national authorities to implement, timely and effectively, the reforms. In 9 particular, it is important to take into account all the external factors beyond the control or influence of the institutions, such as the economic and political environment, administrative capacity, ownership of the reform process, communication policy, and any unexpected shocks (e.g. early elections, a deeper global recession) or other exogenous factors. 110. Regarding financial risks beyond the programme’s horizon, the financing tables and Debt Sustainability Analysis (DSA) of the Commission look at financing conditions after the programme ends. Box 12 The IMF assessment of the fiscal baseline and of fiscal measures has diverged from the assessment of the European institutions in various instances. 123. The implementation of the programmes allowed for preventing the collapse of the financial system with all the dramatic consequences for the Greek state and for financial stability. The origins of the deterioration of the banks' balance sheets lie in the protracted implementation of the financial market reforms under the programme, recurrent periods of political instability with heightened uncertainty and confidence losses, as well as the prolonged recession. 124. Reviving long-term growth is a consequence of successful programme implementation rather than a short-term goal. The success of the programme can only be assessed many years after its end since reforms will continue to have cumulative effects long after a programme is over. CONCLUSIONS AND RECOMMENDATIONS Recommendation 1 The Commission accepts the recommendation. It will specify a framework for establishing conditionality and give more clarity on the types of analytical tools to be used. Recommendation 2 The Commission accepts the recommendation. Recommendation 3 The Commission accepts the recommendation. The Commission notes that the current ESM Treaty provides for more focused programmes, e.g. targeted at imbalances in specific sectors, in which case a comprehensive growth strategy may not be warranted. Recommendation 4 The Commission accepts the recommendation. Recommendation 5 The Commission accepts the recommendation. Most steps in this respect have already been taken as part of the ESM programme. Recommendation 6 The Commission accepts the recommendation and recalls that it cannot commit other institutions to accept working modalities that, by definition, need to be jointly agreed both in principle and in substance. Recommendation 7 The Commission accepts the recommendation. The Commission already applies quality control measures including through cooperation with other institutions. The Commission will review the existing quality controls and improve the related documentation. 10 141. Given the priority of fiscal consolidation, there was a sequencing of fiscal and structural reforms. Further, the real issue was not the design of reforms, but their implementation. Indeed, the administrative capacity was a challenge, therefore technical assistance was provided to Greece on this subject following the establishment of the Commission's Task Force for Greece. Further, the effective use of the technical support provided and actual implementation of structural reforms were hampered by recurrent periods of protracted political instability. As far as quantitative targets are concerned, they were set in cooperation and agreement with the Greek authorities. Recommendation 8 The Commission accepts the recommendation. Technical support has been closely aligned with the provisions of the ESM stability programme, whereby support to a number of reforms under the programme has been explicitly included in the Memorandum of Understanding. Within three months after the ESM programme was established, the Commission agreed with the Greek authorities a "Plan for technical cooperation in support of structural reforms" that was also published on the Commission's website. The Structural Reform Support Service provides and coordinates support to the Greek authorities in almost all reform areas under the ESM programme. Recommendation 9 The Commission accepts the recommendation. It will enhance its analytical work on the design of programme reforms by specifying a framework for establishing conditionality, which will give more clarity on the types of analytical tools to be used. 146. The implementation of the programmes allowed for preventing the collapse of the financial system, with all the dramatic consequences that it would have had for the Greek state, and restoring financial stability which was key to mitigating a further decline. The origins of the deterioration of the banks' balance sheets lie in the protracted implementation of the financial market reforms under the programme, recurrent periods of political instability with heightened uncertainty and confidence losses, as well as the prolonged recession. Recommendation 10 The Commission accepts the recommendation. It has already carried out ex-post evaluations for other Euro area countries which had stability support programmes. Recommendation 11 The Commission accepts the recommendation. 11 Event Date Adoption of Audit Planning Memorandum (APM) / Start of audit 11.11.2015 Official sending of draft report to Commission (or other auditee) 13.7.2017 Adoption of the final report after the adversarial procedure 3.10.2017 Commission’s (or other auditee’s) official replies received in all languages 14.11.2017 We examined the European Commission’s management of the three Economic Adjustment Programmes for Greece, bearing in mind the institutional set-up of the different financial assistance instruments used. In relation to the ongoing programme, the audit focused only on the design aspects. Funding for the first programme (GLF), in 2010, was 110 billion euros; for the second (EFSF; 2012) it was 172.6 billion euros and for the third (ESM; 2015) it was 86 billion euros. As of mid-2017, Greece still requires external financial support and we found that the objectives of the programmes were met only to a limited extent. Overall, the programmes’ design did make the progress of reform in Greece possible, but we found weaknesses. We make a number of recommendations to the Commission for future support programmes. 1977 - 2017 EUROPEAN COURT OF AUDITORS 12, rue Alcide De Gasperi 1615 Luxembourg LUXEMBOURG Tel. +352 4398-1 Enquiries: eca.europa.eu/en/Pages/ContactForm.aspx Website: eca.europa.eu Twitter: @EUAuditors © European Union, 2017 Reproduction is authorised provided the source is acknowledged.