Address https://gelex.eu/eli/EU/doc/http://publications.europa.eu/resource/cellar/d6397b79-e1c3-411f-a793-e5be6f6060dc.0006.01@2016-09-06/en

Contents

1. Preamble
2. (a)
3. (b)
4. ANNEX
5. ANNEX II SOME EXAMPLES OF MISSING INFORMATION IN DOCUMENTS FROM THE SIX MEMBER STATES EXAMINED
6. pt_1
7. ANNEX III SOME EXAMPLES OF CLASSIFICATION WEAKNESSES IN DOCUMENTS FROM THE SIX MEMBER STATES EXAMINED
8. pt_6
9. ANNEX IV SOME EXAMPLES OF ACTION POINTS WITH A POTENTIAL IMPACT ON THE DEFICIT AND DEBT DATA
10. ANNEX V ANNUAL NOMINAL GDP GROWTH RATE
11. ANNEX VI COMPLIANCE WITH THE NOMINAL AND STRUCTURAL TARGETS SET IN EDP RECOMMENDATIONS AND EDP DEADLINE EXTENSIONS
12. ANNEX VII DIFFERENCES BETWEEN COMMISSION AND COUNCIL RECOMMENDATIONS
13. ANNEX VIII IMPLEMENTATION OF PLANNED STRUCTURAL REFORMS

The ECA’s special reports set out the results of its performance and compliance audits of specific budgetary areas or management topics. 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 performance audit was produced by Audit Chamber IV — headed by ECA Member Milan Martin Cvikl — which specialises in auditing revenue, research and internal policies, financial and economic governance and European Union’s institutions and bodies. The audit was led by ECA Member Milan Martin Cvikl, supported by the Head of his private office, Kathrine Henderson and Andreja Rovan, Attaché; Zacharias Kolias, Director; Albano Martins Dias da Silva, Head of Task; Giuseppe Diana, Marko Mrkalj, Maria Isabel Quintela, Marion Schiefele, Duarte Semedo Leite and Stefano Sturaro, auditors; Thomas Everett and Marek Riha, language support. From left to right:G. Diana, M. Schiefele, A. Rovan, A. Martins Dias da Silva, Z. Kolias, M. M. Cvikl, S. Sturaro, M. I. Quintela, K. Henderson, T. Everett. ACR Autonomous Commission recommendation AMECO Macroeconomic database of the European Commission DG Economic and Financial Affairs DirectorateGeneral for Economic and Financial Affairs DG Migration and Home Affairs Directorate General for Migration and Home Affairs ECB European Central Bank ECOFIN Economic and Financial Council edp Excessive deficit procedure EEF European economic forecast EFC Economic and Financial Committee EMU Economic and monetary union EPP Economic partnership programme ESA European System of Accounts Eurostat Statistical Office of the European Union GDP Gross Domestic Product GFS Government finance statistics MTO Mediumterm budgetary objective SCP Stability or convergence programme SDV Standard dialogue visit SGP Stability and Growth Pact TFEU Treaty on the Functioning of the European Union UDV Upstream dialogue visit Actual data Historical (not planned/forecast) data. Bottomup fiscal effort A quantification of the fiscal impact of corrective measures, expressed as the aggregated impact of individual measures. See Topdown fiscal effort. Budget balance The balance between total public expenditure and revenue in a specific year, with a positive balance indicating a surplus and a negative balance indicating a deficit. For the monitoring of Member State budgetary positions, the EU uses general government aggregates. See also Structural (budget) balance##HT_ITALIC_IC#, #HT_ITALIC_IC##Primary (budget) balance. Code of Conduct Policy document with guidelines for the implementation of the Stability and Growth Pact and the format and content of stability and convergence programmes. Convergence programmes Mediumterm budgetary and monetary strategies presented by Member States that have not yet adopted the euro. They are updated annually in accordance with the provisions of the Stability and Growth Pact. See also Stability programmes. Cyclical component of budget balance That part of the change in the budget balance which follows automatically from cyclical elements in the economy, due to the reaction of public revenue and expenditure to changes in the output gap. See Structural (budget) balance. Data quality Data quality means compliance with accounting rules, completeness, reliability, timeliness and consistency (Article 8(1) of Council Regulation (EC) No 479/2009). Dialogue visit Carried out by Eurostat in a Member State to review the actual data reported and review all relevant aspects of EDP data. Member States are visited at least every two years. DG Economic and Financial Affairs and the ECB act as observers. From 2011 to 2013 Eurostat carried out both ‘standard’ and ‘upstream’ dialogue visits – the latter to assess the management and control of upstream public finance/systems data flows. These verifications are now included in standard dialogue visits. Economic and Financial Committee (EFC) Formerly the Monetary Committee, the EFC is a Committee of the Council of the European Union set up by Article 134 of TFEU. Its main task is to prepare and discuss ECOFIN Council decisions with regard to economic and financial matters. EDP inventory of sources and methods Description of the procedures (methods) and basic statistics (sources) used by a Member State to produce estimates for EDP notifications. Includes ‘bridge tables’ showing the correspondence between public accounting items and the ESA headings, and a list of general government units classified by subsector and industry. EDP notifications and related questionnaires EDP notifications Twice a year, Member States send their general government deficit and debt figures in the form of notifications comprising four groups of tables with information, for all levels of government, on the government deficit/surplus and debt levels, the adjustments made to the public accounts working balance to obtain the deficit/surplus, the factors contributing to the level of debt, and other compulsory data. Related questionnaires A set of tables containing standard information on revisions, adjustments, financial transactions and the application of Eurostat decisions. ESA (ESA 95 and ESA 2010) The European system of national and regional accounts is the EU’s internationally compatible accounting framework for the systematic and detailed description of the economies of the EU Member States and regions. For most of the audited period the relevant system was ESA 95, which was upgraded in September 2014 to ESA 2010, reflecting developments in measuring modern economies, advances in methodological research and the needs of users. While the ESA is broadly consistent with the definitions, accounting rules and classifications applied by the comparable UN system, it also has some specific features which are more in line with EU practices. European Semester The annual cycle of economic policy coordination, which takes place over the first six months of the year. The European Commission undertakes a detailed analysis of Member States’ programmes of economic and structural policies, and the European Council and the Council of Ministers provide policy advice before Member States finalise their draft budgets. Financial/nonfinancial corporations Financial corporations sector (ESA 95, paragraph 2.32) Includes all corporations and quasicorporations which are principally engaged in financial intermediation (financial intermediaries) and/or in auxiliary financial activities (financial auxiliaries). Nonfinancial corporations sector (ESA 95, paragraph 2.21) Consists of institutional units whose distributive and financial transactions are distinct from those of their owners, and which are market producers whose principal activity is the production of goods and nonfinancial services. Fiscal consolidation An improvement in the budget balance through measures of discretionary fiscal policy, specified either by the amount of the improvement or the period over which the improvement is sustained. General government ESA 95 defines the ‘general government’ sector as ‘All institutional units which are other nonmarket producers whose output is intended for individual and collective consumption, and mainly financed by compulsory payments made by units belonging to other sectors, and/or all institutional units principally engaged in the redistribution of national income and wealth’. It is composed of four subsectors:

central government - all administrative departments of the state and other central agencies whose responsibilities cover the whole economic territory of a country, except for the administration of social security funds;

state government - institutional units that exercise some government functions below central government and above local government level, except for the administration of social security funds;

local government - all types of public administration whose responsibility covers only a local part of the economic territory, except for local agencies of social security funds;

social security funds - central, state or local institutional units whose main activity is to provide social benefits.

Government debt The total gross debt at nominal value outstanding at the end of the year and consolidated between and within the sectors of general government as defined in ESA 95. It comprises currency and deposits, securities other than shares excluding financial derivatives, and loans (Maastricht debt). Government deficit The level of net borrowing as defined in ESA 95/ESA 2010, with certain adjustments. Government finance statistics Data on the economic activities of general government, including government revenue, expenditure, deficit, transactions in assets and in liabilities, other economic flows and balance sheets. European GFS are produced in accordance with the ESA, with additional interpretation and guidance from Eurostat. Mediumterm budgetary objective (MTO) According to the reformed Stability and Growth Pact, stability programmes and convergence programmes present a mediumterm objective for the budgetary position. The MTO is countryspecific to take into account the diversity of economic and budgetary positions and developments and of fiscal risks to the sustainability of public finances, and is defined in structural terms. See also Structural (budget) balance. Methodological visit An exceptional monitoring and verification visit carried out by Eurostat to a Member State where significant risks or problems with regard to data quality have been clearly identified. Oneoff and temporary measures Government transactions having a transitory budgetary effect that does not lead to a sustained change in the budgetary position. See also Structural (budget) balance. Output gap The difference between actual output and estimated potential output at given point in time. See also Cyclical component of budget balance. Potential GDP The level of real GDP in a given year that is consistent with a stable rate of inflation. If actual output rises above its potential level, constraints on capacity will start to bite and inflationary pressure will build; if output falls below the potential level, resources are lying idle and inflationary pressure will abate. See also Output gap. Primary (budget) balance The budget balance net of interest payments on general government debt. Primary/upstream data Data collected by a national statistical institute (e.g. from administrative records, surveys, census) for the purpose of compiling aggregates (such as EDP transactions). The public accounts of government departments, agencies, etc. are an example of primary data. Public accounts Government financial statements (from departments, agencies and any public corporations classified as part of general government). They record state revenue and expenditure and are the main source for compiling EDP data. Public debt See Government debt. Snowball effect The selfreinforcing effect of public debt accumulation or decumulation arising from a positive or negative differential between the interest rate paid on public debt and the growth rate of the national economy. Stability and Growth Pact (SGP) Approved in 1997 and reformed in 2005 and 2011, the SGP clarifies the provisions of the Maastricht Treaty regarding the surveillance of Member State budgetary policies and the monitoring of budget deficits during the third phase of economic and monetary union. The SGP consists of two Council regulations setting out legally binding provisions to be followed by the European institutions and the Member States, and two resolutions of the European Council in Amsterdam (June 1997). Stability programmes Mediumterm budgetary strategies presented by Member States that have adopted the euro. They are updated annually in accordance with the provisions of the Stability and Growth Pact. See also Convergence programmes. Stockflow adjustment The stockflow adjustment (also known as the debtdeficit adjustment) ensures consistency between net borrowing (flow) and the variation in the stock of gross debt. It includes the accumulation of financial assets, changes in the value of debt denominated in foreign currency, and remaining statistical adjustments. Structural (budget) balance The actual budget balance net of the cyclical component and oneoff and other temporary measures. The structural balance gives a measure of the underlying trend in the budget balance. Topdown fiscal effort A quantification of the fiscal impact of government policy, obtained by looking at the overall change in the structural balance. It may differ from the bottomup quantification of effort due to the incomplete coverage of the latter, secondorder economic effects or inconsistency in the ‘no policy change’ assumption. Working balance Public accounts revenue minus expenditure. The Treaty on the Functioning of the European Union establishes it as a basic rule of budgetary policy that Member States must avoid excessive government deficits. Accordingly, where the reference values for deficit and debt are exceeded a corrective mechanism may be initiated against the Member State concerned. This mechanism, the excessive deficit procedure, is a key element of the EU’s economic governance framework. The Commission’s role in implementing the excessive deficit procedure is to check the quality of the data notified by each Member State, to assess whether the reference thresholds have been breached, or risk being breached, and on this basis to address opinions and recommendations to the Council to act accordingly. The Council then decides, according to the Treaty provisions, whether or not to adopt the Commission recommendations. Member States placed under an excessive deficit procedure are given recommendations to remedy the situation, including a deadline by which to correct the situation, a path for correction and an annual fiscal effort to be delivered. The Commission monitors the Member State’s implementation of corrective measures and reports its findings to the Council. The Council, on the basis of Commission proposals, takes further action as appropriate (lifting the procedure, extending the deadline, setting new targets or imposing sanctions). The Court examined the Commission’s implementation of the excessive deficit procedure between 2008 and 2015, focusing on six Member States. We considered the Commission’s quality assessment of Member States’ EDP data, the quality of its own forecast data and models and whether its assessments were in tune with the Council’s decisions to launch a procedure. Lastly, we examined how the Commission monitored the corrective action taken by Member States under an excessive deficit procedure. We found that, although detailed procedures and guidelines exist for most areas of the Commission’s data collection and analysis and its assessment of compliance with the rules on budgetary discipline, there are problems with its implementation of these tasks. This is because the Commission did not make full use of its powers to enforce the provision of comprehensive data and compliance with recommendations for corrective action, and because of the failure at times to give adequate feedback on Member States’ reports, the insufficient resources devoted to analysing and reporting on key data, and poor recordkeeping. A related issue is that of transparency, where, despite improvements in recent years, too much information is still generally unavailable regarding the Commission’s data assumptions and parameters and its understanding of key concepts. In addition, even where the Commission has set clear internal rules it may decide to depart from the established procedure, which raises questions about the overall reliability of its assessments. In the area of monitoring structural reforms, an aspect of corrective action which the Commission recently highlighted as crucial to overcoming an excessive deficit situation, there are signs that the Commission does not go far enough, as it essentially focuses on legislative aspects rather than the actual implementation of reforms. The excessive deficit procedure continues to overemphasise the criterion of deficit rather than debt. In conclusion, we find very positive signs in the Commission’s efforts over the years to adapt and rationalise the excessive deficit procedure in response to developments in the EU. The legislative basis is sound and is generally supported by clear internal rules and guidelines. What has been lacking is consistency and transparency in the application of those rules; the Commission does not adequately record its underlying assumptions or share its surveillance findings for the greater benefit of all Member States. In recent statements, the Commission has acknowledged these shortcomings and indicated that it is prepared to make the necessary improvements. Our overall conclusions are explained in detail at the end of the report, and a number of specific recommendations are addressed to the Commission. Those recommendations are summarised below.

(a) In the area of data collection and analysis, which is Eurostat’s field of competence, the Commission should:

(a)

(i) enhance its quality assessment procedures and better document its work;

(ii) assess Member States’ own control systems;

(iii) take steps to improve the effectiveness of its onthespot verifications;

(iv) make full use of the powers available to ensure that Member States implement followup action points;

(v) maximise transparency by making public all advice and guidance to Member States;

(vi) better document its internal procedures and criteria for setting reservations or making amendments to data.

(i) improve transparency by applying clear definitions, disclosing all calculation and assessment data and promoting the involvement of national fiscal councils to confirm the national data used in the Commission’s analyses;

(ii) focus closely on the reduction of government debt, especially in heavilyindebted Member States;

(iii) strengthen its monitoring of the implementation of agreed structural reforms, including making full use of its powers to ensure Member States meet their commitments;

(iv) more strictly enforce the rules on Member States’ reporting;

(v) exploit the option, where appropriate, of recommending that the Council take tougher action and impose sanctions.

(b) In assessing the situation in Member States leading to recommendations in connection with the EDP (this is the role of DG Economic and Financial Affairs), the Commission should:

(b)

(i) improve transparency by applying clear definitions, disclosing all calculation and assessment data and promoting the involvement of national fiscal councils to confirm the national data used in the Commission’s analyses;

(ii) focus closely on the reduction of government debt, especially in heavilyindebted Member States;

(iii) strengthen its monitoring of the implementation of agreed structural reforms, including making full use of its powers to ensure Member States meet their commitments;

(iv) more strictly enforce the rules on Member States’ reporting;

(v) exploit the option, where appropriate, of recommending that the Council take tougher action and impose sanctions.

The Treaty on European Union, also known as the Maastricht Treaty, which entered into force on 1 November 1993, established the principles of economic and monetary union (EMU) and led ultimately to the euro as a common currency in 19 out of 28 Member States. The proper functioning of EMU required the introduction of a mechanism to safeguard the soundness of public finances and to reduce the risk of spillover from Member States pursuing unsuitable fiscal policies. This mechanism comprises two arms, one preventive and the other corrective. The preventive arm aims to ensure sound budgetary policies in the mediumterm and avoid excessive deficit situations. The corrective arm addresses excessive deficit situations and is known as the excessive deficit procedure (EDP). Member States that breach the ceiling on budgetary deficit or debt (respectively 3 % and 60 % of GDP) may be subject to an EDP. The primary legal basis for the EDP is Article 126 of the Treaty on the Functioning of the European Union (TFEU) and Protocol No 12 annexed to that Treaty. These specify the different steps in the procedure and the reference thresholds. They are complemented by secondary legislation in the form of the Stability and Growth Pact (SGP), which was introduced in 1997 and provides operational clarifications on the circumstances in which an EDP should be launched. The EDP has undergone several reforms (in 2005, 2011 and 2013) aimed at improving its implementation by strengthening reporting requirements and sanctions and introducing the flexibility to adapt the procedure to a changing economic environment. Figure 1 provides an overview of the evolution of economic and fiscal governance policy in the EU, with measures related to the EDP highlighted in bold. Annex I contains a chronological list of EU legal instruments that are most relevant to the EDP. Note:Measures related to the EDP are highlighted in bold. Source: ECA. Most recently, on 21 October 2015 the Commission adopted a communication on steps towards completing economic and monetary union to implement the ‘Five Presidents‘ Report’. The communication and its accompanying proposals are intended to take forward the process of deepening the EMU and boost competitiveness. The package of measures includes improved tools for economic governance, including the creation of national competitiveness boards and an advisory European Fiscal Board. The Commission has indicated that the steps proposed are both ambitious and pragmatic and will need concerted action from all actors involved. The main parties to an EDP are the Member State, the Commission (DG Economic and Financial Affairs and Eurostat) and the Council.

(i) The Member State is required (a) to report statistical data to Eurostat twice a year (before 1 April and 1 October), (b) to correct the excessive deficit situation by the deadline given by the Council, (c) to deliver the fiscal effort required in the respective EDP recommendation, and (d) to report on action taken.

(ii) Eurostat is responsible for assessing the quality both of actual data reported by Member States and of the underlying government sector accounts.

(iii) DG Economic and Financial Affairs is responsible for assessing the Member State’s compliance with the debt and deficit thresholds and monitoring its action to correct the excessive deficit situation.

(iv) The Council, on the Commission’s proposal, decides to launch, to step up or to abrogate the EDP.

Figure 2 shows how these parties cooperate and interact at the various stages of the EDP. Source: ECA. This section covers the successive stages of the process leading to a decision. Member States are required to compile EDP data in accordance with the European System of Accounts (ESA) and the Manual on Government Deficit and Debt (MGDD), as well as Eurostat’s decisions and guidance notes. These data, reported in standardised ‘notification’ tables and related questionnaires, are sent to Eurostat together with other government finance statistics (GFS) data and an inventory of sources and methods. Each Member State decides how the EDP data are to be compiled. The national statistical institute (NSI) takes the lead role and is responsible for compiling actual data, while planned data are compiled by the ministry of finance. However, the ministry of finance and other government departments and the central bank usually also provide primary data and, in some cases, are jointly responsible for the compilation process. The NSI supplements its estimates from other data sources (financial statements of public units, business surveys, etc.). Eurostat examines and assesses the quality both of the actual data reported by Member States and of the underlying general government sector accounts. Quality data presupposes compliance with the accounting rules (ESA and MGDD), completeness, reliability, timeliness and consistency. The assessment focuses on areas specified in the EDP inventories, such as the delimitation of the general government sector population (including exhaustiveness checks that all public units are covered), the classification of specific government transactions, liabilities and the accrual recording of transactions. Eurostat also makes ‘dialogue visits’ to Member States to verify and assess the EDP data and reporting processes, including whether they comply with the accounting rules. Where significant risks and problems in the quality of data are identified, Eurostat may carry out ‘methodological visits’ (so far only one such visit has been made, to Greece). Following its quality assessment, Eurostat forwards the data to DG Economic and Financial Affairs for economic analysis. However, where there is evidence that the data reported by a Member State do not comply with the accounting rules, Eurostat may express a reservation as to the quality of the actual data or make amendments. It may also initiate an investigation into manipulation of statistics. In such cases fines of up to 0.2 % of GDP can be imposed. DG Economic and Financial Affairs analyses the actual Member State data provided by Eurostat, and its own European economic forecasts produced in the context of the European Semester, to monitor the evolution of deficit and debt and to assess Member States’ plans for convergence towards the reference thresholds. The data are analysed according to a framework of rules and guidance. DG Economic and Financial Affairs maintains regular contact with the Member States throughout this process to obtain information on local developments. The objective of the analysis is to:

(i) assess whether a Member State has breached the deficit and/or debt criteria, taking account of any particular circumstances (severe economic downturn, recent trends and forecasts for indicators, adjustment for potential output, etc.);

(ii) if a breach of either the deficit or the debt criteria is identified in a Member State, prepare a report under Article 126(3), which considers in detail a series of aggravating and mitigating factors, and assesses the case for launching an EDP; the report is sent to the Economic and Financial Committee (EFC) for an opinion; and

(iii) if an excessive deficit situation is recognised, inform the Member State and the Council and recommend that the latter launch an EDP with deadlines for the Member State to take corrective action.

It is the Council which decides whether to place a Member State under an EDP. The Member State is then informed accordingly and the decision is published. The Council sets a deadline for addressing the situation (in principle one year from the detection of the excessive deficit, although several multiannual EDPs were launched in 2009 due to the financial and economic crisis). Table 1 summarises all the Council’s EDP decisions to 17 November 2015. In the case of multiannual EDPs, the Council also sets annual fiscal targets. Through its recommendations, moreover, it usually calls for the development and implementation of national structural measures, such as public expenditure reform. At this point, euroarea Member States that have already been sanctioned for noncompliance with the recommendations based on Article 121(4) TFEU on the preventive arm of the Pact, or whose breach of the deficit and debt ratios is especially serious, may also face a fine in the form of a redeemable noninterestbearing deposit of 0.2 % of GDP. Source: DG Economic and Financial Affairs website (as at 17.11.2015). Closed Ongoing Council decision initiating an EDP Council decision abrogating the EDP Council decision on initiating an EDP Current deadline for correction Belgium 2.12.2009 2015 20.6.2014 2015 - - Bulgaria 13.7.2010 22.6.2012 - - Czech Republic 5.7.2004 2.12.2009 3.6.2008 20.6.2014 - - Denmark 13.7.2010 20.6.2014 - - Germany 21.1.2003 2.12.2009 5.6.2007 22.6.2012 - - Ireland - - 27.4.2009 2015 Greece 5.7.2004 5.6.2007 27.4.2009 2016 Spain - - 27.4.2009 2016 France 3.6.2003 30.1.2007 27.4.2009 2017 Croatia - - 21.1.2014 2016 Italy 28.7.2005 2.12.2009 3.6.2008 21.6.2013 - - Cyprus 5.7.2004 11.7.2006 13.7.2010 2016 Latvia 7.7.2009 21.6.2013 - - Hungary 5.7.2004 21.6.2013 - - Lithuania 7.7.2009 21.6.2013 - - Luxembourg 2010 2010 - - Malta 5.7.2004 2009 7.7.2009 21.6.2013 5.6.2007 20094.12.2012 19.6.2015 - - Netherlands 2.6.2004 2.12.2009 7.6.2005 20.6.2014 - - Austria 2.12.2009 20.6.2014 - - Poland 5.7.2004 7.7.2009 8.7.2008 19.6.2015 - - Portugal 5.11.2002 20.9.2005 11.5.2004 3.6.2008 2.12.2009 2015 Romania 7.7.2009 21.6.2013 - - Slovenia - - 2.12.2009 2015 Slovakia 5.7.2004 2.12.2009 3.6.2008 20.6.2014 - - Finland 13.7.2010 12.7.2011 - - United Kingdom 2004 24.1.2006 20049.10.2007 8.7.2008 Financial year 2016/17 This section describes how corrective measures are reported to and monitored by the Commission. Member States are responsible for developing and implementing corrective measures to restore fiscal stability in accordance with the Council’s recommendations. They are required to keep DG Economic and Financial Affairs informed about the measures planned and/or adopted and developments in their fiscal and economic situation. They must submit a report on action taken and disclose EDPrelevant information in their stability or convergence programmes and national reform programmes prepared in the context of the European Semester. Before the ‘Sixpack’ (2011), Member States were not legally obliged to submit a report on action taken. The ‘Twopack’ then introduced additional reporting requirements for euroarea Member States (see Figure 1 and Figure 7) to enhance the Commission’s monitoring and surveillance. DG Economic and Financial Affairs is responsible for examining whether a Member State has complied with the Council’s recommendations, for monitoring the implementation of EDP corrective measures and for proposing further action. DG Economic and Financial Affairs carries out assessments of the action taken by each Member State under an EDP:

(i) as early as three or six months from the launch of the procedure on the basis of the Member State’s report on the action taken in response to the Council recommendations;

(ii) regular monitoring coinciding with each forecast, in the light also of information disclosed by the Member State in its stability/convergence programme and, for euroarea countries, its draft budgetary plan; and

(iii) at the expiry of the deadline, unless the procedure is abrogated.

More specifically, DG Economic and Financial Affairs assesses whether the Member State has announced or taken measures that seem adequate to ensure sufficient progress towards correction of the excessive deficit within the time limits set by the Council. Depending on the assessment made, DG Economic and Financial Affairs will propose different actions to the Council:

(i) where the Commission considers that the Member State has complied with the recommendations and that the EDP requirements are likely to be fulfilled, the procedure will be ‘held in abeyance;

(ii) where there has been effective action but ‘unexpected adverse events with major unfavourable consequences for government finances have occurred’, a revised recommendation may be proposed. The revised recommendation may extend the deadline for correction of the deficit, usually by one year, or may set out new nominal and structural targets linked to a new underlying macroeconomic scenario, without extending the deadline;

(iii) where there has been no effective action and the Member State is in the euro area, the EDP may be ‘stepped up’. The Council will issue a notice inviting the Member State to take specific measures and to submit reports by set deadlines, and if the Member State fails to comply with this notice the Council can impose sanctions. Where applicable, the Council may decide to suspend payments under the Cohesion Fund and, with effect from the 2014-2020 programming period, the European Structural Investment Funds.

The expected outcome of the EDP decision and corrective action is that the Member State will restore budgetary discipline, and hence the sustainability of its public finances, by correcting the excessive deficit within a reasonable deadline. The EDP decision is also intended to foster the implementation of structural reforms in the Member State, which is conducive to public finance stability in the medium/long term. The Court assessed whether the Commission is managing the implementation of the EDP appropriately. For this purpose we looked at the procedure as applied to Cyprus, the Czech Republic, France, Germany, Italy and Malta between April 2009 and May 2015. In relation to the quality of data, we looked at the period 2008 - 2014 and, where necessary, back to 2005. However, although we analysed some of the details underlying the definition of the EDP threshold values, we did not assess the appropriateness of these values, nor did we assess the quality of the measures taken by Member States to implement the procedure. More specifically, we examined:

(i) the Commission’s assessment of the quality of Member States’ statistical data (Eurostat) and its interpretation of those data (DG Economic and Financial Affairs) leading ultimately to recommendations that the Council initiate or take further action in relation to an EDP;

(ii) the Commission’s monitoring and surveillance tasks; and

(iii) the outcome of the procedure.

The main audit criteria were derived from:

(i) regulatory requirements (e.g. Council regulations, decisions, directives and conclusions); and

(ii) the Commission’s internal rules and procedures (e.g. instructions and guidelines).

We reviewed the documentation relating to the separate EDPs. We also interviewed Eurostat and DG Economic and Financial Affairs staff and representatives of the institutions involved in implementing the procedure in each of the Member States selected for the audit. Table 2 lists the entities visited in each Member State. Finally, we discussed EDP issues with representatives of several organisations whose views were relevant to the audit (including representatives of the IMF and of the United Kingdom Office for Budget Responsibility), and for more technically complex aspects we made use of specialised internal and external expertise. Source: ECA. Supreme Audit Institution National Statistical Institute Ministry of Finance Fiscal Council Central Bank Others Czech Republic √ √ √ Germany √ √ √ France √ √ Italy √ √ √ √ √ Economic Adviser to the Presidency of the Council of Ministers Cyprus √ √ √ √ √ DG Economic and Financial Affairs resident representative Malta √ √ √ √ This section examines the Commission’s procedure for analysing the data and assessing compliance prior to recommending a decision. Eurostat is responsible for assessing the quality of the statistical data used in the EDP process. Its role in this task has developed significantly since 1993, specifically through the introduction of new tools and new discretionary powers (Figure 3 shows how the main legal framework has evolved since the Maastricht Treaty). Source: ECA. Eurostat has itself made significant further efforts to develop its approach and improve the quality assessment process. Between 2008 and 2014 it introduced many new measures. Most importantly, in 2011, having reviewed its EDP data quality assessment structures and processes, Eurostat made resourcing and procedural changes. For example, upstream visits were initiated to assess the quality of the systems generating primary data. See Figure 4 for further details. Source: ECA, based on Eurostat information. Commission recommendations to launch an EDP are predicated on the results of specific analyses by DG Economic and Financial Affairs. These analyses are based in turn on a set of internal rules, instructions and guidelines that explain how they are to be performed and describe the most relevant underlying concepts. Most of these internal rules, instructions and guidelines can be found in a ‘Vademecum on the Stability and Growth Pact’ (available to the public) and ‘Specifications on the implementation of the Stability and Growth Pact’ (also known as the ‘Code of Conduct’). They address all the most relevant issues. Moreover, given the underlying importance of the rules, instructions and guidelines to the Commission’s proposals on the launch of an EDP, DG Economic and Financial Affairs regularly prepares analytical and operational notes for the EFC in order to explain to Member States its approach to interpreting and implementing the SGP. The Commission’s regular European economic forecasts (spring, autumn and winter) are another important part of its analyses leading to a decision whether or not to propose launching an EDP. The generation of forecasts combines elements at country level – prepared by DG Economic and Financial Affairs’s country desks – with centralised elements produced by the horizontal unit (see Box 1). Overall, about 150 staff from different units within DG Economic and Financial Affairs are involved in the process. Two months before the expected publication date, DG Economic and Financial Affairs prepares a position paper setting out general assumptions. Forecasting is then based on an iterative process with crossverification to ensure consistency between the countrylevel and central forecasts. The country desks each prepare countryspecific macroeconomic and fiscal forecasts. To do this they use a spreadsheet tool based on a centrallymanaged template, as well as additional tools that reflect countryspecific characteristics such as data availability and economic structure. The horizontal unit adds followup calculations based on the projections by the country desks (e.g. the output gap and cyclicallyadjusted fiscal figures) and performs quality checks, which include:

ensuring that a common set of external assumptions is used;

checking numerical and economic consistency between country desk forecasts;

checking the consistency of the country desks’ trade flow projections.

The Commission’s forecasting procedure is complex. However, it is structured in a similar way to that of other international organisations (e.g. IMF, OECD, ECB), as a mix of topdown and bottomup elements. This setup allows country forecasts to be tailored to the countryspecific characteristics without losing central oversight. Furthermore, in terms of overall forecast accuracy, the Commission’s economic forecasts do not deviate significantly from those produced by peer organisations. Compared with those of its peers, the Commission’s country forecasts are more heavily reliant on expert judgement and less on econometric approaches. This poses difficulties when it comes to evaluating and assessing their validity. Some Member State authorities interviewed believe that the forecasting process lacks transparency, in particular with regard to the detailed countryspecific assumptions made for fiscal indicators. If differences arise between the Commission’s projections and those made nationally, the Member States find it impossible to evaluate them because the Commission does not sufficiently disclose the details of its assumptions in its talks with national authorities. This is consistent with our experience when reviewing the forecasts. Discrepancies between the Commission’s forecasts and those produced by the Member States can be explained by different underlying assumptions. However, identifying the assumptions that are responsible for such discrepancies is not an easy task given the multiplicity of spreadsheets and expert opinion. In addition, there is no standard summary (e.g. in the form of a reconciliation table) clearly outlining the differences between the Commission’s and Member States’ estimates for the fiscal indicators that are used as the basis for the Commission’s recommendations. The above should be seen in the light of the rules, which state that ‘the Commission shall make public the methodologies, assumptions and relevant parameters that underpin its macroeconomic and budgetary forecasts’. Nevertheless, the Member State authorities also noted that the Commission’s disclosure of information has significantly improved over the years. Changes in economic knowledge and the structure of the economy call for continuous improvements to the forecasting process. In recent years, the Commission has taken a number of steps to enhance its forecasting. For instance, the main forecasting template has been overhauled, the level of detail for some economic sectors has been increased, and an elaborate new tool for detecting additional data inconsistencies has been introduced. Furthermore, the Commission is developing a pilot forecast model, initially for the euro area as a whole, and countryspecific modules, with the eventual aim of using the model to deliver quality checks that complement its judgementbased forecasts. Further improvement is necessary. The Commission publishes the relevant external forecast assumptions and other technical assumptions, but implicit countryspecific assumptions are not described and were very hard to infer from the spreadsheets. Furthermore, DG Economic and Financial Affairs considers the spreadsheets to be internal documents – not to be shared outside the Commission. Eurostat’s examination of statistical data includes verifications carried out either inhouse (data quality assessments) or onthespot (dialogue visits). These are complemented by guidance provided either bilaterally (at the request of a Member State), or, as a Eurostat initiative, on particularly complex issues. Eurostat carries out an extensive analysis of the statistical data received from Member States and their answers to requests for clarification in the three weeks following each of the two annual EDP notifications (see paragraph 10). Outside this period, Eurostat also carries out inhouse verification of methodological and other datarelated issues, complemented by dialogue visits to Member States. There is still room to document the inhouse verification process more comprehensively. At present it can be difficult to see how individual quality assessments were carried out and, therefore, to understand the reasons for the assessment. It can also be more difficult to demonstrate that assessments are carried out in a consistent way. Box 2 gives some examples of weak documentation. In the six Member States visited, there were cases where the authorities were asked if they applied the necessary criteria but were not required to substantiate their replies. Examples included the recording of privatisations and dividends. One Member State sent data on interest and financial statements, but the assessment made by Eurostat was not well documented. Without further substantiation, one Member State simply confirmed as briefly as possible that interest was calculated on an accrual basis. There is no evidence of any systematic or detailed Eurostat analysis of the information in bridge tables. Eurostat has never provided feedback to Member States on the quality of bridge tables. In addition, the data provided by the Member States are not always complete. This applies to both mandatory and voluntary data, as the latter are also often used in Eurostat assessments. Annex II gives some examples of missing information. Most of the quality assessment process is governed by internal written procedures and guidance setting out how data should be analysed. However, it could be made more explicit how the existing rules and guidelines (ESA and MGDD) are to be applied in one key area: the classification (delimitation) of public institutional units within the general government sector (i.e. which entities/bodies are to be included in the calculation of government debt and deficit). This could improve the coverage and consistency of the assessment (see Annex III). We examined the classification of a number of public institutional units in the six Member States in the sample. We found that in some cases the ESA 95 rules for the classification of units should have been analysed in greater depth. We also found that the six Member States applied different time spans for analysing and classifying public units (see Box 3 for some examples).

cases not fully compliant with ESA 95 but accepted by Eurostat, such as exhaustiveness, different approaches - in classifying units;

the main weaknesses found in the classification of public institutional units;

weaknesses in relation to the procedures and timing for analysing and (re)classifying units within/outside the general government sector.

For more details see Annex III. Eurostat has used two types of visits to Member States to verify the data used in the EDP process. The first type is the standard dialogue visit (SDV), which is used to review data, examine methodological issues and assess the NSI’s compliance with the accounting rules. SDVs are made to each Member State at least once every two years. Secondly, from 2011 to 2013 Eurostat also undertook ‘upstream’ dialogue visits (UDVs) as a means of assessing the management and control of upstream public finance/systems data flows from the various providers of EDP statistical data to the NSI, which uses them to compile deficit and debt statistics. The objective of these visits was to identify, assess and monitor significant risks or problems with a view to taking corrective action. UDVs were incorporated into SDVs as of 2014. Table 3 shows the dates of dialogue visits during the audited period. Source: ECA, based on Eurostat information. 2008 2009 2010 2011 2012 2013 2014 Czech Republic August (SDV) May (SDV) May (UDV) November (SDV) November (SDV) Germany May (SDV) May (SDV) June (UDV) October (UDV) February (SDV) France June (SDV) November (SDV) November (SDV) June (SDV) Italy March (SDV) June (SDV) November (UDV) February (SDV) Cyprus December (SDV) December (SDV) September (SDV) September (SDV) Malta January (SDV) July (SDV) May (SDV) May (SDV) Eurostat uses a risk analysis model to identify the risks associated with the collection and compilation of EDP statistics in a Member State. The risk assessment is carried out once per year and per Member State. It is a wellstructured and comprehensive model that covers all the aspects of the compilation process (see Figure 5). The risk analysis was mainly used to rank the Member States to be visited for a UDV. However, it could have been better documented. For example, we did not always find written explanations for changes from one year to the next in the likelihood and/or the impact of the risks assessed. There was also no evidence that highrisk topics were selected for the visits. Source: ECA, based on Eurostat information. From 2006 to 2014, the average duration of SDVs and UDVs in the six Member States selected was 1.7 and 3.5 days respectively. The number of topics to be covered varied from three to ten for SDVs and from four to seven for UDVs. As a rule, the agenda of a dialogue visit includes five standard points and a number of specially selected issues. In exceptional cases, where the country is visited more frequently or the purpose is to discuss a limited number of issues, the agenda might include fewer items. Given that the team consists of three to four staff, it is unlikely that the time on site is sufficient to cover all the ground in sufficient depth and so data quality issues may not be adequately examined. Eurostat’s assessment of EDP data quality is reliant on Member States’ statistical systems. Consequently, it must have assurance, not only that EDP data are compliant with the ESA standards but also that, notwithstanding their different statistical systems, Member States apply a consistent approach to data compilation. During the period examined, Eurostat visits covered the institutional arrangements in place in Member States, namely the existence of formal agreements binding the national authorities under an EDP. Although recommended, this is not a legal obligation. In the course of its UDVs, Eurostat also reviewed the quality of EDP systems, mainly in relation to upstream data. Eurostat does not have formal written rules in this regard, and we did not find any evidence that desk officers assess Member States’ own supervisory and control systems. Without much documentation on the supervisory and control systems in place for data compilation, it is not possible to implement a systemsbased approach for examining the quality of the data received. For the Member States examined, reports were published between nine and 52 weeks after the corresponding visits (see Figure 6). Onthespot verifications may trigger ‘action points’ that the Member State should implement so as to address weaknesses found in the course of the verifications. Action points are nonmandatory and often take a long time to resolve. In the six Member States, approximately one third of action points had not been resolved within the deadline set by Eurostat. However, we found no cases in the sample where failure to respond to the action points within the deadline was taken into account in the quality assessment of the data submitted, even though they sometimes cover important issues which could have an impact on the data. Source: ECA, based on Eurostat information. To ensure a harmonised approach, Eurostat provides methodological guidance to Member States on how specific issues of general government should be recorded. Guidance may be triggered in one of two ways:

(a) Member States may ask Eurostat for advice on complex issues.

(b) During its quality assessment work Eurostat may come across methodological issues requiring interpretation. If these issues concern multiple Member States it sets up a taskforce or organises expert consultation of the Committee on Monetary, Financial and Balance of Payments Statistics.

Eurostat provides bilateral guidance at service level if so requested by a Member State. On particularly complex issues, it also sometimes takes the initiative to provide guidance. Although binding, however, bilateral guidance is not always reported to the EFC and not always made public due to the Member States’ desire for confidentiality. Guidance provided to three of the six Member States (in the form of exante/expost advice) was neither reported to the EFC nor published on the Eurostat website. In 2014 and 2015, Eurostat received 36 formal requests for exante/expost advice, of which only 23 were made public. The publication of such guidance could help with the sharing of practices among Member States and improve comparability and transparency. Some key aspects of the assessment process are unclear, for example the identification of oneoff and temporary measures and the quantification of their estimated impact. These measures may make the difference when deciding whether or not to place a Member State under an EDP, or whether to extend the deadline for correcting an excessive deficit situation (see also paragraph 80). The definition of such measures is vague and therefore open to interpretation. They are defined as ‘measures having a transitory budgetary effect that does not lead to a sustained change in the intertemporal budgetary position’, or ‘nonrecurrent measures whose impact on the general government balance is concentrated in one or a very limited number of years’, but there is no detailed methodology for identifying them. The indicative list in Box 4 is taken from the ‘Public finances in EMU 2006’ report and gives some examples of measures considered to be oneoff and temporary. Tax amnesties implying a oneoff tax payment. The typical case is that of a government offering an amnesty in order to repatriate capital from abroad. The amnesty may be subject to a oneoff tax payment, which can potentially be large. Sales of nonfinancial assets (real estate, publicly owned licenses and concessions). The most famous example is the sale of UMTS licences in 2000 (and in some countries in subsequent years). Temporary legislative changes in the timing of outlays or revenues with a positive impact on the general government balance. This includes changes in tax rates that are clearly announced as temporary and temporary changes in the timing of expenditure and collection of revenues. Exceptional revenues linked to the transfer of pension obligations. These revenues correspond to the payment by a given company to the government, in exchange for the transfer of the responsibility for the future payment of pensions of its employees. The magnitude of the oneoff payment depends on the value of the pension commitments assumed by government and is potentially important. Changes in revenues or expenditure consecutive to Court or other authorities rulings. Such measures include for instance reimbursements of subsidies to general government decided by the Commission. Securitisation operations with a positive impact on the general government balance. Exceptional revenues from State owned companies. Shortterm emergency costs associated with major natural catastrophes or other exceptional events (e.g. military actions, others). Experience shows that the exceptionality of these costs depends considerably on the size of the country. Changes in revenues or expenditure consecutive to Court rulings or consecutive to Commission decisions. Another key aspect requiring clarification is the issue of ‘unexpected adverse economic events with major unfavourable consequences for government finances’ (see paragraph 24). Although a definition of ‘major unfavourable consequences for government finances’ exists for internal purposes, the concept of ‘unexpected adverse economic events’ has never been defined. While it is not possible to exhaustively define the set of individual country circumstances to which the concept may apply, given particularly that such events are by nature unanticipated, criteria are needed with a view to providing some guidance. Faced with the lack of a definition and guidance, when looking at France in 2013 the Commission classified unexpectedly low economic growth as an ‘unexpected adverse event’, but it is difficult to see how this can be justified. This section considers the transparency, objectivity and practicability of the Commission’s reservations and recommendations. Eurostat may express reservations or make amendments if it judges that data quality is compromised. During the period 2008-2014, Eurostat set twelve reservations and amended data in ten cases, none of which concerned any of the Member States examined by us. Considering the large number of action points, Eurostat sets reservations rather infrequently. Eurostat has not explicitly laid down criteria for setting reservations or amending Member States’ data, and the procedure is based on professional judgement. In some cases where Eurostat identified methodological problems, the reasons for not expressing a reservation were not well documented, including the potential impact on the deficit and debt data. Box 5 describes some of the cases concerned.

application of the market/nonmarket test was infrequent and a large number of public units (corporations) were not tested.

amount received by a government unit for assuming the costs of decommissioning nuclear facilities (1.6 billion (0.1 % of GDP) for 2004).

incompleteness of the local government sector.

nonmarket units are not classified in the general government sector from the first year of their nonmarket status.

unavailability of detailed data for the analysis of capital injections, financial derivatives and distributions.

classification of Malita and City Gate project.

recording of gross fixed capital formation is not made on an accrual basis for large investment amounts.

For more details see Annex IV. When an EDP is launched, DG Economic and Financial Affairs produces analyses, which are used as the basis for subsequent EDP decisions, and issues recommendations. Where a procedure is ongoing, it does this following its regular assessments of effective action (see paragraph 22). The results of these analyses are presented, for new procedures, in the report required by Article 126(3) TFEU or, for ongoing procedures, in an internal Commission working document. An EDP will be triggered, in line with the provisions of the TFEU, by the Commission’s examination of Member State compliance with budgetary discipline on the basis of the deficit and debt criteria. To this end, DG Economic and Financial Affairs monitors the development of the budgetary situation and of the stock of government debt in each Member State on the basis of actual data, the plans outlined in the Member State’s stability or convergence programme, and the Commission’s forecasts. A breach of the thresholds laid down in Protocol No 12 to the TFEU does not automatically trigger an EDP, since other relevant factors need to be taken into account. These consist of an openended list of economic indicators grouped under the headings of mediumterm economic, budgetary and government debt positions, as well as any other factors considered relevant and put forward by the Member State. The Article 126(3) reports reviewed covered all the topics mentioned above. The debt criterion only became operational in 2012, is still being phased in and will not apply to all Member States until at least 2020. EDPs initiated before November 2011 are to be abrogated on the basis of the deficit criterion only, and this will be followed by a threeyear transitional arrangement for debt. Box 6 shows how the debt criterion has evolved since the Maastricht Treaty. Protocol No 12 to the Treaty on European Union, signed in Maastricht in 1992, set a threshold level, for the public deficit, of 3 % of GDP and, for debt, of 60 % of GDP or a ratio that is ‘sufficiently diminishing and approaching the reference value at a satisfactory pace’. Assuming a 5 % annual nominal GDP growth rate, a deficit of 3 % of GDP will eventually stabilise the debttoGDP ratio at 60 %. If the ratio is above 60 %, a 3 % deficit will induce an annual reduction in the debt ratio of one twentieth of the differential with respect to the reference value of 60 %. In other words, focusing on the deficit was deemed sufficient to ensure the steady reduction of debt, and for many years it was felt unnecessary to define more precisely the pace of convergence with the reference value. Since the 2008 crisis, the 5 % assumption for annual nominal GDP growth has clearly been unrealistic (see Annex V). Hence a deficit of 3 % of GDP no longer automatically implies convergence towards a debttoGDP ratio of 60 % (for example, at an annual nominal growth rate of 3 % and a 3 % deficit, the debttoGDP ratio will stabilise at 100 %). For this reason the debt criterion was adjusted. A more precise definition emerged in Article 1(2)(b) of Council Regulation (EU) No 1177/2011 of 8 November 2011: ‘the [debttoGDP ratio] shall be considered sufficiently diminishing and approaching the reference value at a satisfactory pace […] if the differential with respect to the reference value has decreased over the previous three years at an average rate of one twentieth per year as a benchmark […]’. However, the 23 Member States already under an EDP in November 2011 are subject to a transitional rule for a period of three years after abrogation of the respective EDP. As a result, the new debt criterion will not apply universally to all Member States before at least 2020 (see also Table 1). The rule since 2011 has been that the government debttoGDP ratio complies with the debt criterion if it does not exceed the reference value (60 % of GDP), or if the differential with respect to the reference value decreases over three years at an average rate of one twentieth per year, on a backward, forwardlooking or cyclically adjusted basis. The acceptable pace of reduction in the debttoGDP ratio is approximately that which would apply, regardless of the initial level of debt, given a public deficit of 3 % of GDP and annual nominal growth of 5 %. As nominal growth has consistently been below 5 % since 2008, to meet the debt benchmark the public deficit must be brought below 3 % of GDP – and when higher levels of initial debt are factored in the deficit target will be more demanding still. This makes the convergence requirements particularly challenging for heavilyindebted Member States. In Table 4 we use the Commission data and projection for the debttoGDP ratio from 2013 to 2020 to calculate the annual ratio resulting from compliance with the debt benchmark for five countries in our sample. Only Germany and Malta, where debt hovers around 70 % of GDP, are shown as meeting the debt criterion over the period under consideration. Cyprus, which was placed under a programme of financial assistance in April 2013, is forecast to be compliant with the debt benchmark in 2018. France, which was placed under an EDP in April 2009 and is expected to exit the procedure by the end of 2017, should be compliant with the debt criterion at the end of a threeyear transitional period starting at the abrogation of the EDP (i.e.in 2020). However, with a debttoGDP ratio of 96.5 % in 2015, France is not expected to fulfil the debt criterion by 2020. As with France, for all years in the period the Commission also projects that Italy’s debttoGDP ratio will exceed the maximum acceptable for compliance with the benchmark. Assuming the scenario projected by the Commission, therefore, Italy, which had a debttoGDP ratio of 133.0 % in 2015, will be unable to comply with the debt benchmark by the forecast horizon of 2020 – and is thus prima facie in breach of the debt criterion. Source: ECA simulation based on EC data – the cells highlighted in yellow indicate that for that year the Member State is forecast to comply with the debt benchmark. 2013 2014 2015 2016 2017 2018 2019 2020 GERMANY debttoGDP (EC) 77.4 74.9 71.4 68.5 65.6 63.0 60.6 58.2 debttoGDP (benchmark) 73.1 70.4 67.6 65.1 62.7 FRANCE debttoGDP (EC) 92.3 95.6 96.5 97.1 97.4 97.8 97.8 97.4 debttoGDP (benchmark) 91.5 92.9 93.4 93.8 94.0 ITALY debttoGDP (EC) 128.8 132.3 133.0 132.2 130.0 128.2 126.0 123.5 debttoGDP (benchmark) 124.5 125.5 124.7 123.3 121.4 CYPRUS debttoGDP (EC) 111.5 117.4 126.0 122.5 116.4 111.1 106.5 102.6 debttoGDP (benchmark) 112.9 116.1 115.5 111.0 106.2 MALTA debttoGDP (EC) 69.6 68.3 65.9 63.2 61.0 59.2 58.1 57.4 debttoGDP (benchmark) 67.1 65.2 63.0 61.0 59.4 The number of relevant factors (see paragraph 14(ii) that the Commission can take into account in its analysis is potentially unlimited. Since, moreover, there is no methodological framework in place to weigh the impact of those factors in the assessment, all conclusions are left to the Commission’s economic judgement. Although certain cases may require a more qualitative approach based on expert knowledge, it is difficult to understand how the Commission’s capacity for economic judgement applies to other factors. In four of the nine Article 126(3) reports examined, the Commission concluded that the relevant factors present or seem to present a ‘mixed picture’, and in three other reports it concluded that the relevant factors seem to present a favourable, or relatively favourable, situation. However, the reasoning behind these conclusions is unclear, as is the impact of this analysis on the final decision whether or not to place a Member State under an EDP or extend the EDP deadline. Information on the data used by the Commission in its calculations was often not easily accessible. This finding was corroborated by the representatives of the national authorities of some of the Member States visited. In the case of revised recommendations, for example, the information in the Commission’s relevant staff working document was not sufficient to replicate the calculation of the fiscal effort and other variables that are essential for the assessment of Member States’ compliance with the original recommendations. In this regard, the working document underlying the Commission’s recommendation to extend the EDP deadline for France in February 2015 did not contain data in support of the baseline scenario or any detail of additional discretionary revenue measures, although these data are essential to replicate the assessment of effective action which led to the extension. We reviewed the Commission analyses underlying the Council’s EDP decisions and recommendations in respect of the Member States in the sample. The legal framework governing the issue of EDP decisions and recommendations provides room for a relatively high degree of flexibility of interpretation and application. This is true for the assessment of the Italian budgetary situation in the Article 126(3) report of 27.2.2015, which concluded that the debt criterion should be considered satisfied. Box 7 summarises our analysis of the case. Italy was placed under an EDP in November 2009. The procedure was lifted in May 2013. Since then, Italy has been required to comply with the threeyear transitional rule for the debt criterion. In February 2015 the Commission carried out an assessment and concluded that the Member State was prima facie in breach of the debt criterion. However, after taking account of the relevant factors, and in spite of the fact that the debttoGDP ratio was still increasing, the Commission assessed Italy to be compliant with the debt criterion. The decision cited three main relevant factors:

(i) The currently unfavourable economic situation, characterised by very low inflation, which make compliance with the debt rule particularly demanding.

Our analysis: the debt adjustment benchmark already takes into account the influence of the economic cycle on the pace of debt reduction. Rules for doing so are set out in a note addressed to the Economic and Financial Committee alternates entitled ‘Operationalising the debt criterion in the Excessive Deficit Procedure’. This document makes it clear that the adjustment of debt for the cycle also factors in the evolution of prices, which is captured by the GDP price deflator.

(ii) Adherence to the mediumterm budgetary objective (MTO) or the path of adjustment towards it. According to DG Economic and Financial Affairs, strict compliance or convergence with the MTO implies, under normal macroeconomic circumstances, compliance with the debt criterion in the medium term and debt sustainability.

Our analysis: based on the 2015 winter EEF, Italy complied with the MTO in 2014 (in the sense that, its economy being in recession, it was not required to make any adjustment towards the MTO – the structural balance actually deteriorated by 0.2 %) and was forecast to comply in 2015. However, although the 2015 winter EEF also covers 2016, the assessment only takes account of 2014 and 2015. According to the Commission’s forecast Italy will not be in line with the path of adjustment to the MTO in 2016 (forecast structural effort -0.2 % instead of the +0.5 % required). Hence, Italy was not forecast to meet the condition of ‘strictly respecting the required adjustment towards the MTO’.

(iii) Structural reforms: in a January 2015 communication, the Commission clarified that the relevant factors to be taken into account when deciding not to launch an EDP include major structural reforms having direct longterm positive budgetary effects and planned in sufficient detail by the Member State. In DG Economic and Financial Affairs’s view, Italy had made progress with the implementation of an extensive reform agenda with measures ‘expected to have a positive impact on growth and therefore on the sustainability of public finance’.

Our analysis: the communication does not quantify the terms ‘major positive impact on growth’ and ‘longterm positive budgetary impact’. Italy itself did quantify the expected impact of reforms on growth, but not the longterm budgetary impact. In addition, the Commission did not share some of the assumptions underlying Italy’s estimate of a positive impact on growth, merely stating that ‘these results have not been endorsed by any national independent institution and seem to overestimate the impact of the reforms’. Finally, there is no methodological framework in place that explains how the budgetary cost of reforms should be estimated and factored into the assessment of compliance with the debt criterion. We focused on Italy because, of the Member States in the sample, it was the only one where an EDP was not launched even though the country failed to meet the debt benchmark; the decision was shaped by considerations drawn from the analysis of other relevant factors. Of the five other Member States, the Czech Republic’s debttoGDP ratio is below the threshold, and Germany is currently meeting the debt benchmark. Malta was under an EDP from 2013, when it missed the debt requirement, until 2015, when it too became compliant for both deficit and debt. The rules regarding surveillance are different for Cyprus, since it was placed under a financial assistance programme in 2013. Finally, in the case of France, which has been under an EDP since 2009, according to Regulation EC No 1467/97 the debt criterion will only apply when the Member State exits the EDP, i.e. not before 2018. In the case of Italy, the Commission took advantage of the option of invoking relevant factors in order to avoid applying the EDP to a Member State which would be highly unlikely to comply with the debt benchmark. In other words, rather than addressing the root causes that made compliance with the benchmark very demanding (because, for example, it does not take account of the initial level of debt), the Commission put off the problem by emphasising relevant factors which made strict compliance with the benchmark a matter of secondary importance. For three of the six Member States in the sample the EDP deadline was extended and revised recommendations were issued (see Table 5). Source: ECA. Original recommendations Revised recommendations Date Deadline Average annual fiscal effort Commission assessment date Date of recommendation Extension Deadline Average annual fiscal effort France April 2009 2012 At least 1 % of GDP starting in 2010 November 2009 November 2009 1 year 2013 Above 1 % of GDP over the period 2010-2013 May 2013 June 2013 2 years 2015 1.3 % of GDP in 2013, 0.8 % of GDP in 2014, 0.8 % of GDP in 2015 February 2015 March 2015 2 years 2017 at the latest 0.5 % of GDP in 2015, 0.8 % of GDP in 2016, 0.9 % of GDP in 2017 Cyprus July 2010 2012 At least 1.5 % of GDP over the period 2011-2012 May 2013 May 2013 4 years 2016 Not specified Malta July 2009 2010 Not specified February 2010 February 2010 1 year 2011 0.75 % of GDP in 2011 We consider that the analysis underlying the March 2015 extension of the French EDP, based on 2015 winter forecast, was not fully transparent. In particular, the revised recommendation was based on a bottomup assessment which stated that, ‘overall the available evidence does not allow concluding that the recommended effort has not been delivered in 2013 – 2014’. This despite the fact that the calculation of structural effort included the effects of a discretionary revenue measure which, being temporary (financial impact from 2014 to 2016), should not have been taken into account. The measure in question was worth 0.1 % of GDP. Without it the structural effort would have fallen short of the 1.1 % needed to justify extension of the EDP deadline. Moreover, the Commission regarded France as compliant in retrospect and invoked compliance to justify extending the deadline so as to bring the deficit below the threshold at a more relaxed pace. However, this assessment was based only on 2013 and 2014 data. The forecasts available for 2015, which showed that France was going to miss its targets, were not taken into account in the Commission’s assessment (see Annex VI). According to the Vademecum on the SGP, the assessment of effective action should be brought forward if there is a risk that the outturn data at the expiry of the deadline will show that the headline deficit target has been missed. The purpose of this recommendation is to prevent any conflict between Article 10(3) of Regulation EC No 1467/97, which requires the EDP to be stepped up immediately if, according to the outturn data, the headline target is missed, and Article 3(5) of the same regulation, which allows for an extension of the deadline on the basis that effective action was taken (i.e. the Member State achieved the required fiscal effort as assessed using topdown and/or bottomup indicators). However, according to the information available at the time of the assessment France was not on course to meet any nominal or structural indicators by the end-2015 deadline. Rather than extending the deadline, therefore, the Commission should have either issued an early warning (through an autonomous recommendation) or stepped up the procedure. The Commission made use of the high degree of flexibility and discretion afforded by the SGP rules in its assessments of Italy and France. We compared EDP decisions and recommendations in the version finally adopted by the Council with those proposed by the Commission, and found a number of differences for which the available records did not give underlying reasons. For example, in the case of Cyprus (06/07/2010) the Commission recommended 1.75 % of GDP over the period 2010-2012. This was changed in the Council’s recommendation to 1.5 % of GDP over the period 2011-2012. For the most relevant discrepancies see Annex VII. The ‘Sixpack’ and the ‘Twopack’ introduced new reporting requirements with the aim of enhancing surveillance of the EDP (see Figure 7). However, there is no statutory arrangement for ensuring that Member States comply with these reporting requirements. As a consequence, the Commission does not take any action against Member States which fail to provide the required information. Furthermore, our analysis showed that the Commission did not bring this issue to the attention of the European Parliament and the Council in its November 2014 review of economic governance and the application of the ‘SixPack’ and ‘TwoPack’. The Member States in our sample that are subject to the new reporting requirements have not always complied with them (see Figure 8 for examples). Source: ECA. Source: ECA. With the exception of Cyprus (after it agreed to a financial assistance programme in 2013), corrective measures were not sufficiently detailed and/or their fiscal impact was not sufficiently clearly identified or explained (see Box 8) for the Commission to exercise effective supervision. Moreover, the deadlines set for the production of European economic forecasts are not aligned with national budgetary deadlines. This results in frequent discrepancies between the Commission forecasts and national budget forecasts, and consequently between budgetary targets and the required EDP adjustment paths.

The Czech national reform programme (2012) contained a list of measures, both planned and implemented, as well as implemented measures with an expected financial impact in 2013 and 2014. However, the programme did not include any estimate of the fiscal impact of these measures.

The Czech convergence programme (2012) included a table specifying the fiscal impact of measures envisaged for 2013-2015. However, the data in the text could not be reconciled with the information in the table. Moreover, in one case there were no details on the corrective measures proposed.

The Maltese report on action taken (2013) specified corrective measures for 2013 but was vague about 2014. Instead of giving details, the report merely referred to the forthcoming budget. Therefore the Commission’s assessment did not incorporate the planned measures (they were not included until the winter forecast), which weakened the basis for effective monitoring and surveillance.

The French stability programmes for 2010 to 2014 did not describe the corrective measures for every year in the requisite detail. As a consequence, the Commission did not have a sufficient basis for effective supervision or decisionmaking.

This section describes the rules and procedure for assessing effective action, examines the quality of the Commission’s assessments and considers the impact of recent changes. The rules for assessing effective action have evolved as a result of:

(i) the SGP reforms in 2005 and 2011, which complemented the initial deficit benchmark (3 % - SGP 1997) with a topdown approach based on the structural balance; and

(ii) additional interpretative guidance, endorsed by the Council in 2014, including a bottomup assessment methodology.

Compliance with the EDP recommendations is checked by assessing whether, on the basis of the actual data and Commission forecasts, the Member State meets the annual deficit targets. If the Member State fails to do so, an assessment of effective action is carried out using the topdown and bottomup approaches (see Box 9). If the annual deficit targets are not reached, the topdown approach is used to assess whether the Member State achieved the improvement in the structural balance (‘recommended fiscal effort’ R) required in the EDP recommendations. This is done by comparing R with the observed or forecast change in the structural balance (‘observed fiscal effort’ ΔS) and with the change in the structural balance adjusted for the impact of (a) revisions in potential output growth, (b) revisions in the composition of the economic growth or other windfalls/shortfalls on revenues, and (c) other unexpected events (‘adjusted fiscal effort’ ΔS*). If both ΔS and ΔS* are equal to or greater than R, the Member State is considered to have taken effective action and the EDP is held in abeyance. In all other cases, the assessment needs to be confirmed by a careful analysis, which is carried out using the topdown and bottomup approaches. The bottomup approach aims at assessing the fiscal effort by providing a direct estimate of the budgetary impact of new fiscal measures implemented by the Member State since the EDP recommendation was issued, or since compliance with the EDP recommendation was last assessed. The fiscal effort is calculated as the sum of (a) the estimated budgetary impact of additional discretionary revenue measures and (b) the change in nominal expenditure compared to a baseline scenario which the Commission projected earlier on the assumption of no policy change. Introduction of the topdown and bottomup approaches has made the rules for assessing effective action more complex. Topdown approach: the structural effort depends on an estimate of the output gap, which is understood as the difference between actual and potential output. Potential output is defined as the maximum level of output that can be obtained without generating inflationary pressure. It is an unobservable variable and estimating it is fraught with uncertainty (see Box 10). Since the 2005 reform, potential output has played a crucial role in the EDP. This theoretical concept is unobservable, even in retrospect, and has to be estimated. There are various methods of estimating potential output. Despite the view of experts that ‘there is significant uncertainty around any estimate of the output gap’ and the fact that ‘it would be unwise to base an assessment of economic prospects on any single approach alone’, the ECOFIN Council did adopt such a single approach: the production function methodology. This methodology uses a predefined functional form for the production function and estimates potential output by combining, after removing their cyclical component, the capital stock, the labour force and the total productivity of factors. The same production function applies to all Member States: a CobbDouglas function with an output elasticity of labour equal to 0.65 and an output elasticity of capital equal to 0.35. These values were obtained using ‘the mean wage share for the EU15 over the period 1960-2003 […] as the estimate for the output elasticity of labour, which gives a value of 0.63 […] and, by definition, 0.37 for the output elasticity of capital.’ However, ‘since these values are close to the conventional mean values of 0.65 and 0.35, the latter are imposed for all countries.’ Revisions of potential output estimates are triggered by the adjustment of past data and the use of new data on output and expected output, but also by other technical issues and underlying hypotheses – such as expected changes in inflation. This is why potential output estimates are subject to frequent and possibly significant revisions which may also affect past estimates. Actual output figures (both historical and forecast) may also be substantially revised. The result of all this is that the output gap is often revised and may fluctuate enormously – in some cases the magnitude of revision is greater than the previous estimate of the output gap itself. Figure 9 illustrates this volatility. For example, the 2013 output gap in Cyprus was initially estimated in the 2011 forecast at -1.32 %; it fell to -6.16 % in the spring 2013 forecast and rose to -4.1 % in the spring 2014 forecast. Note: A=autumn, S=spring, W=winter Source: Commission’s annual macroeconomic database (AMECO). The bottomup approach is also somewhat complex, owing mainly to the difficulty in defining the impact of revenue measures and to the ‘no policy change’ benchmark against which government actions will be assessed. The Commission’s scrutiny/certification of the estimated budgetary impact of fiscal measures inevitably depends on information and figures provided by the national authorities. It is impossible for the Commission to adequately assess the accuracy of these figures with the resources it makes available for the task. The increased complexity of the rules for assessing effective action has widened the Commission’s scope for interpretation and discretion. As a result its analyses have become less transparent. With regard to the topdown approach, it is not possible to recalculate the structural effort without access to certain key data which were not shared in full with the Member States until the 2015 spring forecast and have still not been made public (see paragraphs 74 and 75). The Commission used the bottomup approach for France in May 2013. This methodology enabled the Commission to judge that France had taken effective action and to extend its EDP deadline. However, the methodology was not endorsed by the Council until June 2014. In addition, the impact assessment of revenue measures is a projection, not an expost assessment, and therefore may be overly optimistic compared with the actual outcomes. Moreover, the impact of expenditure measures is assessed against a baseline identified by the Commission under the ‘no policy change’ assumption and is a matter for expert judgement. The Commission has been making efforts to enhance transparency. In part since the 2014 spring forecast, and in full since the 2015 spring forecast, it has shared with Member States the information they need to replicate its effective action calculation. However, as most of this detailed information is not made public, an interested third party would not be able to replicate the assessment. In common with the EDP reporting documents submitted by national authorities (see paragraph 86), the Commission assessments reviewed by us contained no detailed (measurebymeasure) or quantified expost assessment of the corrective measures implemented by Member States. Since the entry into force of the ‘Sixpack’, Member States placed under an EDP or granted an extension have been required to produce a report, within the following six months, on the corrective action taken (see Figure 7 in paragraph 85). This concerned two Member States in the sample (Malta and France in October 2013). In both cases, we found that the conclusions of the Commission’s assessment were based on incomplete data. In particular:

(i) the Commission pointed out that Malta (EDP deadline end of 2014) failed to meet the EDP requirements for 2014 in both nominal and structural terms, and that neither the report on action taken nor the draft budgetary plan provided sufficient detail about the discretionary measures underpinning the budgetary targets;

(ii) the Commission’s topdown assessment showed that, in France (extended EDP deadline end of 2015), the structural effort was insufficient for both 2013 and 2014, but the adjusted structural effort was on target. In accordance with the rules, a bottomup analysis was also carried out to confirm whether the Member State had taken effective action. However, the second analysis showed that for both years the structural effort fell short of the Council recommendations (i.e. the budgetary impact of new discretionary measures was insufficient). As for 2015, the last year of the EDP period, the Commission warned that ‘a significant set of measures on top of those already specified will be needed to ensure that the target for 2015 is reached’. As a matter of fact, France missed that target.

Notwithstanding these comments on both Malta and France, the Commission’s conclusion for both reports on action taken was that the Member State ‘has taken effective action and no further steps in the procedure are needed at present’. The Commission argued that the assessment of the report on action taken is retrospective. However, this is not in line with the rules set out in the Vademecum on the SGP (paragraph 232 ‘…the Commission examines [the Member State’s report on action taken] to see whether the Member State has complied with the Article 126(7) recommendations. This is done by assessing whether the Member State is forecast to meet all the nominal targets, according to the Commission forecasts’), and it also means that setting unambitious targets for the first year of an EDP (when the report falls due) could turn the assessment of the report on action taken into a mere formality that provides no added value. In this respect we note, for example, that the Council recommendations issued in June 2013 required France that year to achieve a headline deficit equal to the forecast made for 2013 in the EEF underlying the EDP recommendations, and required only those discretionary measures which the Member State had already approved in its most recent budget at that time. Lastly, the requisite structural effort was simply derived from the headline budget target. The assessment of effective action should allow the procedure to be stepped up if the Member State is missing, or is forecast to miss, all the targets (headline deficit, structural effort, adjusted structural effort and bottomup targets). However, we found that this was not the case with the ongoing assessments (‘regular monitoring’) that were made of France on the occasion of the 2014 spring and autumn forecasts: although France was missing all its targets, the procedure was not stepped up (see Annex VI). Our analysis of the assessment that the Commission carries out at the expiry of the deadline for revised recommendations (Malta in 2010 and France in 2009, 2013 and 2015) is given in paragraphs 79 et seq. On the issue of debt, we found in general that, even after the application of the debt criterion, the analysis of the dynamics of the debt ratio is inconclusive for Member States under an EDP that was already in place in November 2011. The analysis broadly describes historical and forecast trends in the debttoGDP ratio, and warns that Member States’ projections might underestimate the debt growth rate and that measures planned in their stability programmes could be insufficient to bring the debttoGDP ratio down. However, the conclusions do not even question whether or not the debt criterion was breached. The Commission pointed out that any such conclusion would only become relevant after the abrogation of the EDP. In this respect we also note that, according to the internal version of the Vademecum on SGP, EDPs that were open in November 2011 are to be abrogated on the basis of the deficit criterion only. In principle, an EDP should be closed by the end of the year following that in which it was opened. The targets set by EDP recommendations therefore have an annual horizon. However, in 2009 several EDPs were launched with a multiannual deadline. A general government deficit target was set for the period as a whole, with, to this end, a fiscal effort averaged over the period (see Annex VI, yellow cells). The drawback was that this allowed Member States to postpone most corrective action until the last year of the period. At that point, the sheer magnitude of the corrections needed made them unrealistic and forced the Council to extend the deadline. To remedy this weakness, since 2011 EDP recommendations have included annual fiscal targets for these multiannual EDPs in both nominal and structural terms, thus restoring the annual benchmark for assessing corrective action as was the case before 2009. Excessive public deficits may be rooted – at least in part – in structural weaknesses. Structural reforms address those weaknesses and remove obstacles to the fundamental drivers of growth by, for example, liberalising labour, product and service markets, which in turn encourages job creation and investment and improves productivity. Thus structural reforms may improve competitiveness, boost growth and, in the medium and long term, contribute to a lasting correction of the deficit. Since May 2013 the Commission has also taken structural reforms into account in the overall assessment of EDP action. It does this by assessing economic partnership programmes (EPPs). Since the entry into force of the ‘Twopack’, euroarea Member States under an EDP have been required, together with their reports on action taken, to submit EPPs, which are a sort of road map of planned structural reforms. Moreover, since February 2015 the Commission has regarded structural reforms planned in sufficient detail as a relevant factor for not launching an EDP or for granting a multiannual deadline. EPP assessments are a oneoff exercise, and the Commission does not systematically follow up the weaknesses identified. For example, in its opinion on the EPP submitted by Malta in October 2013 the Council concluded that all reforms were a work in progress and that implementation remained subject to risks. In addition, some countryspecific recommendations had not yet been fully addressed. It therefore invited Malta to provide additional information on the implementation of the planned reforms in the upcoming national reform programme and stability programme, while considering additional measures to ensure longterm public finance sustainability. However, although Malta failed to address all the aspects of the Council’s Opinion in respect of a number of reforms, the Commission did not mention the missing details in its assessment of the relevant national reform programme. With the exception of Member States under a financial assistance programme (e.g. Cyprus), followup of the implementation of structural reforms by Member States under an EDP is done only in the context of the European Semester (monitoring of the implementation of countryspecific recommendations) and is no stricter than that carried out in respect of other Member States. Moreover, the monitoring of countryspecific recommendations does not always extend to the actual implementation of reforms or to assessing their effectiveness. It mainly focuses on the steps taken to put the proposed measures into law (see Box 11). We consider that weak monitoring could compromise the impact of the Commission’s recent decision to take greater account of the implementation of structural reforms. By way of illustration of the Commission’s approach to the assessment of structural reforms, reference is made to three reforms regarded as priorities by both Italy and the Commission:

Spending review: in their 2013 and 2014 national reform programmes the Italian authorities highlighted the appointment of a commissioner to oversee the spending review. However, the commissioner appointed in October 2013 resigned in October 2014 (two years before the end of his mandate) after putting forward a spending review plan. The Commission in its assessment (COM (2015)113 final) only noted that ‘in the context of the spending review a choice for more political ownership has been made but only some instruments are fully operational’. It did not assess Italy’s progress (or lack of it) in implementing the commissioner’s proposals. A new spending review commissioner was appointed only five months later.

Tackling corruption: DG Economic and Financial Affairs’s assessment of the anticorruption authority failed to detect any weaknesses undermining its effectiveness. These weaknesses only came to light in June 2014 after a report by DG Migration and Home Affairs had identified several issues (February 2014) and a major corruption scandal broke in Italy (EXPO 2015 in Milan in April 2014). Two other major corruption cases emerged later the same year (the Mose project in Venice in June, Mafia Capitale in Rome in December).

Tackling tax evasion: in its assessment the Commission did not examine the figures disclosed by Italy as the outcome of measures implemented under this heading in the stability and national reform programmes.

This section summarises what has been done to further improve budgetary discipline and examines the effectiveness of the EDP as a corrective and structural reform mechanism. Recent reforms to the EU fiscal framework (the ‘Sixpack’ in 2011 and ‘Twopack’ in 2013 – see Figure 1) were made with the aim of achieving and maintaining the soundness of public finances. In the context of the EDP, the reforms introduced complementary rules, an earlywarning mechanism and a range of new monitoring and surveillance tools; they also strengthened sanctions and eased the conditions for imposing them. The budget balance and debttoGDP ratio are the two Treaty indicators used to assess the budgetary position of a Member State (see paragraph 1). The 2005 reform of the SGP supplemented these by introducing the structural balance as one of the targets that Member States under the EDP must achieve in order to correct an excessive deficit situation. Of the 23 Member States under an EDP in November 2011, 15 had exited the procedure as of the end of 2015 (see Table 1). All of the Member States we examined improved both their budget balance and their structural balance during the audited period (see Figure 10). The Czech Republic, Germany, Italy and Malta managed to exit the procedure. Cyprus agreed to a programme of financial assistance, and France remains under the EDP after three extensions to the deadline. France is the only Member State in our sample with a deficit that still exceeded the 3 % threshold in 2015. Source: ECA, based on AMECO (last update 5 May 2015). The debttoGDP ratio of all the Member States in the sample was higher in 2015 than in 2008 (see Figure 11). In mitigation, it must be recognised that the EDPs we examined were introduced in the midst of a severe financial crisis which contributed to a rapid increase in the debt level through a variety of channels, for example capital injections to banks and low or negative nominal growth. On average, the debttoGDP ratio in the EU increased by more than 24 % of GDP between 2008 and 2012, and all countries were affected. Source: ECA, based on AMECO (last update 5 May 2015). In Germany, Malta and the Czech Republic (where it remains under the threshold), the increase over the 2008-2015 period was moderate (4.4 percentage points in Malta, 6.4 percentage points in Germany and 12.8 percentage points in the Czech Republic) and the ratio is now on a downward path. Cyprus has had to cope with an unprecedented banking crisis and experienced an increase of 61.4 percentage points in its debttoGDP ratio from 2008 to 2015, with the sharpest rise coming between 2008 and 2013 (from 45.3 % to 102.2 %). However, since 2013, when it was placed under a programme of financial assistance, the debt dynamics have eased, and the ratio stabilised in 2014 (107.5 %) and 2015 (106.7 %). Finally, both, France and Italy experienced a continuous increase of around 30 percentage points in their debttoGDP ratio over the 2008-2015 period. The drivers behind debt dynamics are the primary balance, the snowball effect and the stockflow adjustment. The evolution of the debttoGDP ratio can therefore be broken down by the respective impact of those three drivers. Figure 12 shows that breakdown, and Box 12 provides a deeper analysis of the breakdown and of the debt dynamics for the Member States in the sample. The analysis in Box 12 suggests that the EDP framework is potentially effective at keeping debt at bay. However, a high initial level of debt (as in Italy) may hamper the effectiveness of the procedure in keeping the debttoGDP ratio under control, as may the reluctance of a Member State (e.g. France) to implement the EDP recommendations and ensure fiscal discipline. Italian debt continued to increase, in spite of efforts towards fiscal discipline (as illustrated by a positive annual primary balance of +1.1 % on average over the period), owing to a significant snowball effect resulting from a combination of a large initial stock of debt, low or negative nominal GDP growth and episodes of high interest rates. France’s debttoGDP ratio increased despite a moderate initial level due to a lack of budgetary discipline, as illustrated by insufficient improvement in the budgetary indicators (see Figure 10) and by a significant negative primary balance (-2.7 % on average over the period). Source: ECA, based on AMECO (last update 5 May 2015). At the outset of the 2008 financial crisis, Italy’s public debt was 102.3 % of GDP. From 2009 to 2015, Italy had to cope with a difficult economic situation, which was reflected in a strong negative output gap (-3.3 % on average) and a yearly average nominal growth rate of 0.0 %. Although Italy improved its structural balance from 2009 (see Figure 10), the debttoGDP ratio rose by 30.8 percentage points. The debt dynamics were mainly driven by a strong snowball effect, since the primary surplus broadly offset the stockflow adjustment (see Figure 12). The snowball effect can be explained by the high stock of debt, the weakness of nominal growth and the impact of the sovereign debt crisis on interest rates. France’s debttoGDP ratio was 68.1 % in 2008. During the 2009-2015 period France faced a less severe economic turndown than Italy, with annual nominal growth of +1.3 % and an average output gap of -1.6 %. Despite a lower initial level of debt and better cyclical conditions, France experienced a comparable increase in the debttoGDP ratio (28.3 percentage points) during the period. The increase was mainly due (around 2/3) to a lack of budgetary discipline: during the period France ran an annual primary deficit of 2.7 %. The final 1/3 of the increase can be explained by the snowball effect. In Germany the debttoGDP ratio stood at 65.1 % in 2008. Over the 2009-2015 period Germany benefited from a higher nominal annual growth rate (2.4 %) than France and Italy and a less negative output gap (-1.1 % on average) and experienced a rise of 6.4 percentage points in its debttoGDP ratio. This can be explained by better cyclical conditions combined with a protracted period of fiscal consolidation. However, the primary surplus only partially offset the stockflow adjustments that were the main driver of the increase in debt - the snowball effect also playing a limited role. In 2008 Cyprus had a debttoGDP ratio of 45.3 %. Over the period Cyprus faced a deep crisis, with negative average nominal growth (-1.0 %) and an output gap of -2.1 %. The increase of 61.4 percentage points (from 45.3 % in 2008 to 106.7 % in 2015) in its debttoGDP ratio can be explained by the negative impact of all the drivers: primary balance (-18.4 percentage points), snowball effect (-26.4 percentage points) and stockflow adjustment (-16.6 percentage points). The Czech Republic’s debttoGDP ratio increased from 28.7 % to 41.5 % between 2008 and 2015, and thus remained below the threshold. In an economic environment characterised by average nominal growth of 1.5 % and an output gap of -1.5 %, the increase in debt (+12.8 percentage points) can mostly be explained by the primary deficit (-12.5 percentage points), with the stockflow adjustment offsetting the snowball effect. In 2008, Malta had a debttoGDP ratio of 62.7 %. Over the 2009-2015 period the country experienced average nominal growth of 4.6 % and an average nominal deficit of -2.8 %. As a result its debttoGDP ratio essentially remained stable, with an increase of 4.4 percentage points that can mostly be explained by the stockflow adjustment. The EDP provides for the possibility of sanctions against euroarea Member States for poor budgetary discipline. This can be done by stepping up the procedure whenever the Commission assessment shows that a Member State has failed to take effective action to correct its excessive deficit in time. Sanctions could also mean the imposition of a fine of up to 0.5 % of GDP. However, since its inception, the EDP has been stepped up only once, for Belgium,and no sanctions have been applied so far. In our view this entails the risk that Member States will believe that it is unlikely that sanctions will be imposed. Such a perception would undermine the effectiveness and credibility of the EDP. Indeed, although the imposition of sanctions is not the real aim, a system devoid of sanctions is one that relies on nothing more than moral suasion, in which case, unless the Commission can win the Member States’ cooperation, the EDP is bound to be ineffective. Structural reforms are neither binding nor enforceable; therefore Member States are likely to make them subject to their governments’ political agenda. The Commission is unable to influence or boost their implementation. As a result, most of the Member States in the sample only made partial or limited progress towards their reform objectives (see Annex VIII). The outcome of EDP monitoring of the implementation of structural reforms often simply consisted in repeating the Council’s countryspecific recommendations from year to year. Of the countries in our sample, only Cyprus made substantial progress in implementing structural reforms once it agreed to a financial assistance programme (April 2013). The latest available assessment shows that Cyprus has implemented the majority of the agreed measures in full, albeit with some delay. The programme placed Cyprus under closer and more comprehensive monitoring and stricter enforcement rules. When presenting its proposals to implement the ‘Five Presidents’ Report’ (see paragraph 4), the Commission acknowledged that the framework of European economic governance ‘has deepened and widened in scope over the past years, but has also gained in complexity’, and that a first review of the strengthened framework had ‘identified some areas for improvement, notably concerning transparency, complexity and predictability of policy making, which are relevant to the effectiveness of the tools’. The Commission then expressed a commitment to pursuing ‘the full and transparent application of the available instruments and tools’, and to improving clarity and reducing complexity, with the ultimate aim of improving the effectiveness of the existing rules. We consider the proposals a positive development and encourage the Commission to abide resolutely by its commitments. The Commission has taken significant steps towards rationalising and simplifying its use of the excessive deficit procedure as a tool for better economic governance throughout the EU, in a context of increasing complexity. Most recently, it has stated a commitment to change in areas where we identified problems during the audited period: the need for a comprehensive outlook and for greater clarity and transparency to maximise the effectiveness of extensive rules and guidelines. The rules are in place, and implementation of the EDP is on the right track but will benefit greatly from the Commission’s commitment to further improvements. More detailed conclusions on the areas of implementation covered by the audit are set out below, together with specific recommendations. The Commission (Eurostat) has improved its quality assessment processes since 2011 at the level of both resources and procedures. Standardising and increasing the number of requests for information has allowed EDP data to be more extensively analysed. The Court found that the process of examining statistical data is not sufficiently welldocumented for the analysis underlying quality assessments to be easily understood. In addition, it is based on data that are not always complete. It could also be made more explicit how the existing rules and guidelines are to be applied (see paragraphs 47 to 51). The Commission (Eurostat) should better document and enhance the transparency of its work in relation to inhouse verifications. This would give management and other desk officers a clearer overview of the content of analyses. The Commission (Eurostat) should establish internal written procedures guiding the work of analysis in the quality assessment and better assess and document its inhouse verification of data in areas such as the delimitation of the general government sector and specific government transactions. The Commission (Eurostat) does not adequately assess Member States’ control systems in order to ensure that, despite Member States’ having different statistical systems, they apply a consistent approach to the compilation of data. Owing to the general absence of documentation on the supervisory and control systems in place for data compilation, it is not possible to implement a systemsbased approach for examining the quality of the data received (see paragraph 55). The Commission (Eurostat) should enhance its quality assessments by including an examination of the supervisory and control systems in place in Member States, and clearly document its analyses. Written procedures should be drafted and implemented to ensure consistency. Between 2011 and 2013, the Commission (Eurostat) supplemented its ‘standard’ dialogue visits (SDVs) by introducing ‘upstream’ visits (UDVs) to analyse the management and control of upstream data. UDVs proved to be very useful to the quality assessment tasks of identifying, assessing and monitoring risks and problems. In 2014, UDVs were merged with SDVs, with the resulting visits each lasting two days. Given the large number of items on visit agendas, the duration of visits is not sufficient for important data quality issues to be adequately examined (see paragraphs 52 and 54). Eurostat performs limited direct verifications during its onthespot visits. A wellstructured risk analysis has been introduced, but this is not used to preselect highrisk issues for visits. The elements of the risk analysis used to determine which Member States should receive a dialogue visit, and the topics to be discussed during such visits, could be better documented (see paragraph 53). Eurostat’s dialogue visit reports are published anything from nine to 52 weeks after the corresponding visits. Delays in publication mean that Member States are unable to share good practices and may prevent other stakeholders/users from recognising at the appropriate time that certain problems exist in a Member State (see paragraph 56). The Commission (Eurostat) should adapt the duration of its dialogue visits to Member States, or carry out more visits, to ensure more complete coverage. The Commission (Eurostat) should enhance its direct verifications as part of onthespot visits. The Commission (Eurostat) should further reduce the time taken to publish its visit reports. Action points are often not resolved within the recommended deadline. This entails a risk that problems relating to important issues will not be adequately taken into account in Eurostat’s quality assessments (see paragraph 56). The Commission (Eurostat) should make full use of its powers to enforce action in the areas of completeness, methodological responsibility and delivery of documents. When so requested for reasons of confidentiality, Eurostat does not report to the EFC or make public all the exante/expost advice which it provides to Member States on a bilateral basis. This practice compromises transparency and comparability (see paragraphs 57 and 58). To enhance transparency, the Commission (Eurostat) should report to the EFC and make public all advice which it provides on methodological issues. Where confidentiality concerns prevent publication of a document in full, at least a summary should be provided. The reasons for Eurostat reservations and amendments to data could be made more transparent (see paragraph 65). Eurostat has not explicitly laid down written internal procedures for setting reservations as to data quality or making amendments (see paragraph 65). The Commission (Eurostat) should better document its internal procedures and criteria for setting reservations or making amendments to data in order to demonstrate that all methodological problems with a potential/actual impact on net lending/net borrowing are addressed in a consistent way. In general, the analyses underlying the Commission’s proposals for EDP decisions and recommendations lack transparency and, in certain key aspects, are based on vague definitions. Moreover, subsequent reforms of the EDP have increased the complexity of the analytical process. All this makes it difficult in some cases to establish a clear link between the analysis made and the conclusions drawn (see paragraphs 59 to 62, 72 to 83, 91 to 97 and 100 to 104). There is therefore a need to balance the increased complexity and wider scope for economic judgement by enhancing transparency and thus facilitating public scrutiny. In recent years, the Commission has made commendable efforts to improve transparency. However, these efforts are not yet sufficient and need to be pursued further (see paragraph 98). The Commission (DG Economic and Financial Affairs) should further enhance the transparency and clarity of the EDP in the following areas:

Assessment process: the Commission should be more transparent about the key aspects of its analyses. It could achieve this by, for example, devising a methodology to identify and quantify the impact of oneoff and temporary measures, by providing sufficiently detailed explanations of its assessments and conclusions in countryspecific documents - regarding, notably, the concept of ‘unexpected adverse economic events’ when assessing effective action - and by establishing a methodological framework to weigh the impact of the relevant factors that it takes into account in its assessments.

Accessibility of data: the Commission should make public all the detailed information needed to perform the calculations on which its analyses are based, thus enabling any interested third parties to replicate the assessments.

Bottomup assessment: national fiscal councils should be involved in the assessment process by inviting them to provide independent scrutiny of the reliability of the figures and information provided by ministries of finance and used by the Commission in its analyses.

Robustness of analysis: the Commission should consistently make use of all the relevant information made available in its forecasts, abide by the relevant methodologies and disseminate them as widely and promptly as possible.

The EDP has so far focused more on government deficit than on debt. The debt criterion only became operational in 2012, is still being phased in and will not apply to all Member States until at least 2020. The assessment of the debt dynamics of Member States under an EDP initiated before November 2011 is inconclusive: these EDPs are to be abrogated on the basis of the deficit criterion only, and this will be followed by a threeyear transitional arrangement for debt (see paragraphs 68 to 69 and paragraph 104). Flexibility addons to the EDP have meant that, in unfavourable economic times, compliance may be recognised even if the debttoGDP ratio increases. This may undermine the credibility of the process and put at risk the sustainability of public debt, especially during protracted periods of recession (when debt can rise even if the debt criterion is met). Ultimately, whether or not the debt criterion is met, the key thing is to repay or sustainably refinance the actual amount of debt (see paragraphs 76 to 78). The Commission should focus closely on debt developments, especially in heavilyindebted Member States, to prevent debt building to unsustainable levels. Where the 60 % debt threshold is exceeded and the concerned Member State is placed under the EDP, the Commission should ensure that the required adjustments provide for a realistic and credible convergence path towards compliance with the debt rule, in particular by taking into account the initial level of debt. In its EDP recommendations for Member States exceeding the debt threshold, the Commission should, when setting the annual deficit targets, also specify the annual debttoGDP levels that are consistent with those targets. Excessive public deficits may be at least partly due to structural weaknesses. Targeted structural reforms can therefore help to bring about lasting deficit corrections. Since May 2013, structural reforms have been gaining importance as a factor for the Commission’s EDP assessments. However, they are still not enforceable and the Commission is no more diligent in monitoring their implementation than it is for Member States not under an EDP (see paragraphs 108 to 111). Moreover, the monitoring of countryspecific recommendations does not always extend to the actual implementation and effectiveness of reforms. This could compromise the impact of the Commission’s recent decision to take greater account of the implementation of structural reforms (see paragraph 112). In conjunction with the MacroImbalances Procedure, the Commission (DG Economic and Financial Affairs) should make full use of the European Semester in monitoring the implementation of structural reforms for Member States under an EDP, in particular where fiscal structural reforms are included in an economic partnership programme. Besides focusing on legislative aspects, it should seek to assess the effectiveness of the implemented reforms. The Commission should make full use of its powers to ensure that Member States under an EDP meet their commitments with regard to structural reforms. The ‘Sixpack’ and ‘Twopack’ introduced new reporting requirements for Member States with the aim of enhancing the Commission’s surveillance of the EDP. However, these requirements are nonenforceable, in that there is no legislative provision to ensure that Member States comply in this regard, or for sanctions against those that do not. Thus the effectiveness of the Commission’s surveillance is dependent on the active cooperation of the Member States. We found one case where the Member State had not sent reporting documents as required. However, the Commission has not made this public in its corresponding assessments (see paragraphs 85 to 86). Even when it receives full documentation, the Commission does not always assess or follow up the information supplied, which raises concern that it is failing to make best use of the Member States’ reports or that Member States are being burdened with a needless bureaucratic requirement (see paragraphs 99 and 110). The Commission (DG Economic and Financial Affairs) should always make clear in its assessments whether the Member States have fulfilled their reporting requirements. The Commission should make use of the possibility to launch infringement procedures when Member States do not comply with their reporting obligations. Successive EDP reforms have strengthened sanctions against Member States that do not comply with recommendations, and have eased the conditions for imposing them. Since the EDP’s inception, no sanctions have been applied (see paragraphs 114 and 122). While sanctions are also useful as a deterrent, not applying them when Member States fail to fulfil their commitment to budgetary discipline brings the risk that they will be perceived as a tool unlikely to be used. This would undermine their credibility and effectiveness, and hence that of the EDP as a whole. In line with existing legislation, the Commission (DG Economic and Financial Affairs) should recommend that the Council step up the procedure and apply sanctions when there is evidence that a Member State has not complied with EDP recommendations and therefore has failed to fulfil its commitment to budgetary discipline under the Treaty. This Report was adopted by Chamber IV, headed by Mr Milan Martin CVIKL, Member of the Court of Auditors, in Luxembourg at its meeting of 25 February 2016. For the Court of Auditors Vítor Manuel da Silva Caldeira President

ANNEX

Council Regulation (EC) No 3605/93 of 22 November 1993 on the application of the Protocol on the excessive deficit procedure annexed to the Treaty establishing the European Community, amended by:

√ Council Regulation (EC) No 475/2000 of 28 February 2000. Updated the conceptual framework underlying the EDP to ESA 95.

√ Commission Regulation (EC) No 351/2002 of 25 February 2002. Clarified the references to ESA 95 in the context of the EDP by replacing the codes ‘B.9’ and ‘D.41’ with ‘EDP B.9’ and ‘EDP D.41’ respectively.

√ Council Regulation (EC) No 2103/2005 of 12 December 2005, on the quality of statistical data in the context of the EDP.

Council Regulation (EC) No 479/2009 of 25 May 2009 on the application of the Protocol on the excessive deficit procedure annexed to the Treaty establishing the European Community (replaced amended Regulation No 3605/93). Amended by:

√ Council Regulation (EU) No 679/2010 of 26 July 2010 amending Regulation (EC) No 479/2009 as regards the quality of statistical data in the context of the excessive deficit procedure.

Council Regulation (EC) No 1467/97 of 7 July 1997 on speeding up and clarifying the implementation of the excessive deficit procedure, as amended by the Council.

Regulation (EC) No 1056/2005 of 27 June 2005 and Regulation (EU) No 1177/2011 of 8 November 2011 – the corrective arm of the Stability and Growth Pact.

From the ‘Sixpack’:

√ Regulation (EU) No 1177/2011 of the European Parliament and of the Council of 8 November 2011 amending Regulation (EC) No 1467/97 on speeding up and clarifying the implementation of the excessive deficit procedure;

√ Regulation (EU) No 1173/2011 of the European Parliament and of the Council of 16 November 2011 on the effective enforcement of budgetary surveillance in the euro area.

From the ‘Twopack’:

√ Regulation (EU) No 473/2013 of the European Parliament and of the Council of 21 May 2013 on common provisions for monitoring and assessing draft budgetary plans and ensuring the correction of excessive deficit of the Member States in the euro area.

Intergovernmental Treaty on Stability, Coordination and Governance in the Economic and Monetary Union (Fiscal Compact) – 2 March 2012.

ANNEX II SOME EXAMPLES OF MISSING INFORMATION IN DOCUMENTS FROM THE SIX MEMBER STATES EXAMINED

Source: ECA, based on Eurostat information. Missing information in the 2012, 2013 and 2014 April notifications For year N-1: Missing data on trade credits and advances (AF.71 L) in table 4 In one Member State no data was provided until 2009, and in another there was none before 2011. Missing data in EDP questionnaires Five of the six Member States returned EDP questionnaires with incomplete mandatory fields. Examples of incomplete information: Other information During 2008-2014, one Member State did not send the bridge tables1 and the inventory as requested, and three Member States submitted after the deadline. The EDP inventories were due by December 2013, but two Member States sent them after the deadline – one in April and the other in November 2014.

In one Member State:

pt_1

(i) some detailed data for the social security funds subsector and for local government (including the working balance) was missing for entities that are not part of local government and for other adjustments to local government data;

(ii) no data was provided on the contribution to the debt level by longterm loans (increase and reduction) for all levels of general government (April 2012 only) or on the breakdown of shares and other equity for state and local government.

In another Member State, data were not provided for some subsectors (for the working balance, financial transactions included in the working balance and ‘other adjustments’ on state government, local government and social security funds). For all levels of general government, most of the nonmandatory details requested were missing.

(i) four Member States sent incomplete data on capital injections, or none at all, for all years, and one Member State for year N-1;

(ii) three Member States sent an incomplete table 4 (Breakdown of other accounts receivable/payable reported in EDP tables);

(iii) although completeness has improved, two Member States sent incomplete tables 5 (Taxes and social contributions: other accounts receivable/payable of general government), 8 (Central government claims, debt cancellation) and 9 (Guarantees).

ANNEX III SOME EXAMPLES OF CLASSIFICATION WEAKNESSES IN DOCUMENTS FROM THE SIX MEMBER STATES EXAMINED

Source: ECA, based on Eurostat information and documents from the six Member States examined. Main cases not fully compliant with ESA 95 but accepted by Eurostat Main weaknesses in the classification of public institutional units Other weaknesses

in three Member States, the statistical register of public units did not provide exhaustive coverage, mainly at local government level;

one Member State took a prudent approach, that is, automatically allocated to the general government sector – as ancillary units – all public units where sales to general government represented more than 80 % of total revenue, and applied the 50 % criterion to the remaining public units. The prudent approach is an automatic criterion that does not take into consideration the activity of the unit (type of services provided) or other conditions, such as tendering and market. The accounting rules do not specify what is to be done in these cases. Moreover, some units were classified on a group basis (e.g. hospitals) instead of being classified individually according to their market/nonmarket nature;

one Member State took a prudent approach, but only with units below a certain size; for these units the 50 % criterion was applied so as to exclude ‘other revenue’ from sales and include all types of taxes in the cost of production. This was done because the Member State was unable to identify the subsidies to deduct from sales. However, although all taxes were included on the costs side, according to the accounting rules only ‘other taxes on production’ should have been included.

one Member State classified the supervisory unit for financial markets, which is an institutional unit, within the general government sector. According to ESA 95, units of this kind should be classified outside general government (namely, as financial auxiliaries within the financial corporations sector). Other Member States correctly classified these institutional units outside general government;

in one Member State we examined public units that were nonmarket and had been reclassified in October 2014 following the ESA 2010 benchmark revision. The Member State had reclassified a total of 158 public units in the general government sector, 20 because of the new ESA 2010 rules and the remaining 138 as part of a normal benchmark revision. The second group would have been reclassified within the general sector if they had previously been analysed correctly.

In relation to the procedures and timing for analysing and (re)classifying units within/outside the general government sector, we found different approaches in the six Member States:

pt_6

(i) two Member States analyse and (re)classify units every five years under successive benchmark revisions, but monitors new units every year. However, in 2011 one had an action point recommending that the 50 % criterion should not exceed a period of 3-4 years, while the other had an action point recommending the market/nonmarket testing of all public units on an annual basis;

(ii) three Member States analyse units every year and reclassify them if their nature (market/nonmarket) is different for three successive years;

(iii) one Member State analyses units twice a year, or when funding conditions change, and reclassifies them when the nature of the units changes.

ANNEX IV SOME EXAMPLES OF ACTION POINTS WITH A POTENTIAL IMPACT ON THE DEFICIT AND DEBT DATA

Source: ECA, based on Eurostat reports on visits to the six Member States examined. Czech Republic Application of the market/nonmarket test was infrequent and a large number of public units (corporations) were not tested During its visit in 2009, Eurostat found that units with ‘commercial’ legal status (e.g. jointstock companies, limited companies) had not been tested so far and set an action point accordingly. During the 2012 SDV, Eurostat set an action point because the government control test for large companies that are less than 50 % governmentowned was not applied and therefore these units were not analysed. Despite the potential impact no reservation was expressed. Germany Reclassification of nonmarket units to the general government sector from the first year of their nonmarket status In its February 2014 visit, Eurostat detected the problem and set a deadline for compliance of May 2014. The matter was considered closed in July 2014. Clarifications sent in October 2014 showed that the NSI had only applied the action point to units classified after February 2014. Despite the potential impact (not known) no reservation was expressed. Unavailability of detailed data for the analysis of capital injections, financial derivatives and distributions Eurostat first raised the problem of unavailable data following its visit in 2007. The 2009 visit found that the situation had improved for central government swaps and for dividends. Eurostat detected certain problems but did not set an action point, and the matter of dividends was not discussed in detail. Action points were set in 2011 and 2013. In 2011 Eurostat noted that detailed data were unavailable for a proper analysis of capital injections and distributions, and that no local government data were available for swaps, but no action point was set. The 2013 UDV did result in action points, but the problem remains. Despite the potential impact (not known) no reservation was expressed. France Decommissioning of nuclear facilities - Marcoule (1.6 billion (0.1 % of GDP) for 2004) This action point was first discussed at the time of the 2006 visit but, owing to disagreement with the NSI about the correct classification, Eurostat postponed the decision. The case concerned the recording of a lumpsum payment received by a government unit against the assumption of decommissioning costs, with all other obligations to be discharged by the operator. In 2005 Eurostat had given advice on this transaction, proposing that it be classified as prepayments for the purchase of services, but the NSI disagreed and instead recorded a capital transfer, which improved the net lending/net borrowing figures. In 2008 the issue was discussed again but no further action was taken. Despite the known impact no reservation was expressed. Sector delimitation Following its visit in 2012, Eurostat set an action point in connection with the fact that the general government sector was incomplete because it did not include public corporations belonging to/controlled by local government. As of the end of 2014 the NSI had still not provided an exhaustive list. Despite the potential impact (not known) no reservation was expressed. Italy Accrual recording of gross fixed capital formation In its 2012 UDVs Eurostat set an action point in connection with the fact that large investment amounts were recorded on a cash rather than an accrual basis. The NSI was given until 2014 to calculate the correct amounts. In 2013 an action point was set requesting a progress report. The NSI corrected the amounts in the October 2014 EDP notification. Despite the potential impact (not known) no reservation was expressed. Malta Malita and City Gate project Eurostat had detected a problem in its analysis of the revisions in the EDP notification of October 2011. In 2012 it set an action point as a followup to discussions: the NSI was urged in short order to send Eurostat a description of the statistically relevant aspects of the Malita and City Gate project, together with its view of the statistical recording of operations. This point was analysed by Eurostat under bilateral guidance; it was agreed in August 2012 and implemented in October the same year. The impact on the deficit was 0.4 % of GDP. No reservation had been expressed in April 2012.

ANNEX V ANNUAL NOMINAL GDP GROWTH RATE

Source: ECA, based on AMECO database.

Source: ECA, based on AMECO database.

ANNEX VI COMPLIANCE WITH THE NOMINAL AND STRUCTURAL TARGETS SET IN EDP RECOMMENDATIONS AND EDP DEADLINE EXTENSIONS

Annex VI illustrates how the EDPs introduced for the Member States in the sample have evolved. In each case the blue cells show the year of the EDP deadline, with blue arrows pointing to any extensions. Other colours show whether the Member State complied (green) or not (red) with the targets set in the Council’s EDP recommendations for the relevant indicators (headline deficit, structural effort and, for recommendations issued after the ‘Sixpack’ entered into force, the adjusted structural deficit and bottomup indicator). The first number in a cell represents the target, and the second is the actual value achieved. Finally, yellow cells indicate that no annual target was set for the year in question, black cells that the specific indicator was not applicable (not yet in force) and grey cells that, although the data were available, the Commission did not take them into account in its assessment of effective action. Source: ECA, based on the Council’s EDP decisions and recommendations, Commission’s Article 126(3) reports and staff working documents underlying the assessments of SCPs, draft budgetary plans and the 4th, 6th and 7th reviews of the economic adjustment programme for Cyprus.

ANNEX VII DIFFERENCES BETWEEN COMMISSION AND COUNCIL RECOMMENDATIONS

Source: ECA. Commission recommendation Council recommendation Cyprus 6.7.2010 the Cypriot authorities should ensure an average annual fiscal effort of at least 1¾ % of GDP over the period 2010-2012 the Cypriot authorities should ensure an average annual fiscal effort of at least 1½ % of GDP over the period 2011-2012 Czech Republic 30.11.2009 the Council invites the Czech authorities to continue with the necessary pension and healthcare reforms, reforms aimed at raising labour supply and skill levels, and reforms increasing the amount and effectiveness of public R&D the Council invites the Czech authorities to continue with the necessary pension and healthcare reforms France 6.4.2009 the French authorities should stand ready to implement the necessary annual efforts going beyond those foreseen in the subsequent years of the correction period in order to ensure that the annual targets are respected and that the deficit is brought below the reference value by 2012 even if downside risks to the budgetary targets were to materialise the French authorities should implement the necessary efforts to bring the deficit below the reference value by 2012 France 30.11.2009 the French authorities should ensure an average annual structural budgetary adjustment of 1¼ % of GDP over the period 2010-2013 the French authorities should ensure an average annual fiscal effort of above 1 % of GDP over the period 2010-2013 France should improve the overall competition framework, with particular emphasis on network industries, further reform the pension system, modernise employment protection and enhance lifelong learning to improve potential GDP growth France should further reform the pension system as planned which would contribute to longterm fiscal sustainability France 18.6.2013 the French authorities should report on progress made in the implementation of these recommendations at least every six months, as well as in a separate chapter of the stability programmes, until full correction of the excessive deficit has taken place Nil Germany 30.11.2009 the Council invites the German authorities to implement reforms with a view to raising potential GDP growth. This includes (…) reforms to improve the framework for competition in services and promote the integration of the lowskilled and longterm unemployed into the labour market through a flexicurity approach which combines better access to qualifications with improved incentives to work the Council invites the German authorities to implement reforms with a view to raising potential GDP growth Italy 30.11.2009 the Italian authorities should ensure an average annual fiscal effort of 0.5 percentage points of GDP over the period 2010-2012 the Italian authorities should ensure an average annual fiscal effort of at least 0.5 percentage points of GDP over the period 2010-2012 the Italian authorities are also invited to step up efforts to strengthen competition in product and services markets, simplify legislation, reduce the administrative burden at all levels of government and, within a ‘flexicurity’ approach and with a view to reducing regional disparities, improve the functioning of the labour market and the efficiency, outcomes and standards of the education system Nil

ANNEX VIII IMPLEMENTATION OF PLANNED STRUCTURAL REFORMS

2011 2012 2013 France CSR1 Partly implemented Some progress Some progress / limited progress / some progress CSR2 Partly implemented Some progress Some progress / some progress / some progress CSR3 Partly implemented Some progress Some progress / some progress / some progress CSR4 Partly implemented Substantial progress Limited progress/limited progress / some progress CSR5 Partly implemented Limited progress No progress / no progress / some progress Italy CSR1 Partly implemented Some progress Limited progress CSR2 Partly implemented Fully addressed / some progress Limited progress CSR3 Partly implemented Some progress Some progress CSR4 Partly implemented Some progress Limited progress CSR5 Partly implemented Limited progress Limited progress CSR6 Partly implemented Some progress Limited progress Malta CSR1 Partially implemented Not implemented Some progress CSR2 Not implemented No progress Limited progress CSR3 Partially implemented Some progress Some progress CSR4 Not implemented No progress Limited progress CSR5 Partially implemented Limited progress Some progress CSR6 Some progress Germany CSR1 Partially implemented CSR2 Partially implemented CSR3 Partially implemented CSR4 Partially implemented Source: ECA, based on the Commission’s assessments of stability/national reform programmes. 2011 2012 2013 Cyprus CSR1 Partially implemented Partially implemented CSR2 Partially implemented Partially implemented CSR3 Partially implemented Partially implemented CSR4 Partially implemented Partially implemented CSR5 Partially implemented Partially implemented CSR6 Partially implemented Partially implemented CSR7 Partially implemented Partially implemented Czech Republic CSR1 Implemented all / not addressed fully Some progress Some progress CSR2 Partially implemented No progress Limited progress CSR3 Partially implemented Limited progress No progress CSR4 Partially implemented Some progress Limited progress CSR5 Partially implemented Some progress Limited progress CSR6 Partially implemented Limited progress Limited progress CSR7 Limited progress

Versions

1. 2016-09-06