Explanatory Memorandum to COM(2010)1 - Greek government deficit and debt statistics - Main contents
Please note
This page contains a limited version of this dossier in the EU Monitor.
| dossier | COM(2010)1 - Greek government deficit and debt statistics. |
|---|---|
| source | COM(2010)1  |
| date | 08-01-2010 |
Contents
- Brussels, 8.1.2010
- Executive summary
- 1 Introduction 6
- 2 Institutional framework 7
- 2.1 European Statistical System and quality assurance 7
- 2.2 European Statistics Code of Practice 8
- 2.3 Institutional aspects of EDP reporting 8
- 2.4 Governance and institutional framework for EDP reporting in Greece 10
- 3 Chronology of main events 12
- 3.1 The 2004 methodological report 12
- 3.1.1 Exceptional revisions in the September 2004 EDP notifications 12
- 3.1.2 The 2004 methodological report and the infringement procedure 13
- 3.2 Greek reporting between the EDP methodological visit of 29 May-2 June 2006 and 27-29 September 2006, the action plan and follow-up actions (up to 17 July 2008) 13
- 3.3 Greek reporting in 2008: the follow-up EDP methodological visit of 2-3 June 2008 and second EDP methodological visit of 15-19 September 2008 17
- 4 Summary of methodological issues 20
- 4.1 Working balance of the state budget 20
- 4.2 Revision of Treasury Accounts 21
- 4.3 Revenues from abolished extra-budgetary accounts 21
- 4.4 Swaps write-offs 22
- 4.5 Adjustment for interest payments 22
- 4.6 Debt assumptions and guarantees 22
- 4.7 Capital injections 23
- 4.8 Social security funds 24
- 4.9 Local government sub-sector 24
- 4.10 Expenditure on military equipment 24
- 4.11 Recording of tax revenues 25
- 4.12 Extra-budgetary funds and DEKA 25
- 4.13 EU grants 26
- 4.14 Hospital liabilities 26
- 5 Conclusions 28
COM(2010) 1 final
REPORT
ON GREEK GOVERNMENT
DEFICIT AND DEBT STATISTICS
EUROPEAN COMMISSION – JANUARY 2010
Table of Contents
3.2 Greek reporting between the EDP methodological visit of 29 May-2 June 2006 and 27-29 September 2006, the action plan and follow-up actions (up to 17 July 2008) 13
3.3 Greek reporting in 2008: the follow-up EDP methodological visit of 2-3 June 2008 and second EDP methodological visit of 15-19 September 2008 17
3.4 Greek government deficit and debt figures: revisions April 2009 – October 2009 17
Executive Summary
This report responds to the ECOFIN Council conclusion of 10 November 2009 which invited the Commission to prepare a report on the renewed problems in the Greek fiscal statistics. The Council also invited the Commission to propose the appropriate measures to be taken in this situation. The Commission will respond to this latter invitation in the context of the recommendation for a Council decision under Article 126(9) TFEU and an action plan to tackle the Greek statistical problems, including governance.
On 2 and 21 October 2009, the Greek authorities transmitted two different sets of complete Excessive Deficit Procedure (EDP) notification tables to Eurostat, covering the government deficit and debt data for 2005-2008, and a forecast for 2009. In the 21 October notification, the Greek government deficit for 2008 was revised from 5.0% of GDP (the ratio reported by Greece, and published and validated by Eurostat in April 2009) to 7.7% of GDP. At the same time, the Greek authorities also revised the planned deficit ratio for 2009 from 3.7% of GDP (the figure reported in spring) to 12.5% of GDP, reflecting a number of factors (the impact of the economic crisis, budgetary slippages in an electoral year and accounting decisions). According to the appropriate regulations and practices, this report deals with estimates of past data only.1
Revisions of this magnitude in the estimated past government deficit ratios have been extremely rare in other EU Member States, but have taken place for Greece on several occasions. These most recent revisions are an illustration of the lack of quality of the Greek fiscal statistics (and of macroeconomic statistics in general) and show that the progress in the compilation of fiscal statistics in Greece, and the intense scrutiny of the Greek fiscal data by Eurostat since 2004 (including 10 EDP visits and 5 reservations on the notified data), have not sufficed to bring the quality of Greek fiscal data to the level reached by other EU Member States.
Where the EDP notification of 21 October 2009 is concerned, the data have not been validated by Eurostat and a substantial number of un-answered questions and pending issues still remain in some key areas, such as social security funds, hospital arrears, and transactions between government and public enterprises. These questions will need to be resolved, and it cannot be excluded that this will lead to further revisions of Greek government deficit and debt data particularly for 2008, but possibly also for previous years.
As the report shows, the revision in the Greek government deficit statistics are due to two different but in some instances linked sets of problems: problems related to statistical weaknesses and problems related to failures of the relevant Greek institutions in a broad sense. The first set of problems concerns methodological weaknesses and unsatisfactory technical procedures in the Greek statistical institute (NSSG) and in the several other services that provide data and information to the NSSG, in particular the General Accounting Office (GAO) and the Ministry of Finance (MOF). The second set of problems results from inappropriate governance, with poor cooperation and lack of clear responsibilities between several Greek institutions and services responsible for the EDP notifications, diffuse personal responsibilities, ambiguous empowerment of officials, absence of written instruction and documentation, which leave the quality of fiscal statistics subject to political pressures and electoral cycles.
In particular, the report has found evidence of:
- Severe irregularities in the EDP notifications of April and October 2009, including submission of incorrect data, and non-respect of accounting rules and of the timing of the notification;
- Poor cooperation between the national services involved in the compilation of EDP figures, as well as lack of independence of the NSSG and the GAO from the Ministry of Finance;
- An institutional setting and a public accounting system inappropriate for a correct reporting of EDP statistics, especially non-transparent or improperly documented bookkeeping, which has lead to several, and in some cases significant, revisions of data by the Greek authorities over an extended period of time;
- Lack of accountability in the individual provision of figures used in EDP notifications, (e.g. absence of written documentation or certification in some cases, exchange of data by phone);
- Unclear responsibility and/or lack of responsibility of the national services providing source data or compiling statistical data, combined with ambiguous empowerment of officials responsible for the data.
These findings indicate that, on top of the serious problems observed in the functioning of other areas involved in the management of Greek public revenues and expenditures, that are not the object of this report, the current set-up does not guarantee the independence, integrity and accountability of the national statistical authorities. In particular the professional independence of the NSSG from the Ministry of Finance is not assured, which has allowed the reporting of EDP data to be influenced by factors other than the regulatory and legally binding principles for the production of high quality European statistics.
Monitoring instruments available to Eurostat with respect to increasing the transparency of the EDP-related statistics, controlling the quality of data under the quality assurance system and the governance structure, have shown to be subject to shortcomings and limitations of particular relevance in the case of Greece. The problems faced in Greece go well beyond what can be tackled using only the statistical monitoring tools available to the Commission, which according to the Council Regulation 479/2009 does not have audit powers. In spite of the concerted and consistent efforts carried out by Eurostat services since 2004 to ensure respect of the applicable rules and methods, the situation can only be corrected by decisive action of the Greek government.
The Greek authorities need to tackle resolutely not only the outstanding methodological issues, but also and crucially they need to put in place transparent and reliable working practices between the national services concerned, and to revise the institutional setting in order to guarantee the professional independence and full accountability of the NSSG and of the other services involved in the domain of EDP data. Unless the institutional weaknesses identified in this report are addressed and proper checks and balances introduced, the reliability of Greek deficit and debt data will remain in question.
The Commission is fully committed to continue cooperating with the Greek authorities with a view to supporting their efforts to improve the collection and processing of government statistics in order to address the existing shortcomings and restore confidence in Greek statistics.
1Introduction
On 22 October 2009, Eurostat published its biannual press release on government deficit and debt data for the Member States, euro area (EA16) and EU27, covering 2005-2008 data. In it, it expressed a reservation (in line with Council Regulation 479/2009 on the application of the Protocol on the excessive deficit procedure) on the quality of the Greek data 'due to significant uncertainties over the figures notified by the Greek statistical authorities'.
The reservation concerned mainly 2008 data, due to the considerable revision (compared to average revisions in EU Member States) of the deficit figure made between the April 2009 and October 2009 EDP notifications. In fact, the Greek authorities reported twice in October: on 2 October and 21 October, with data in the second notification including a significant revision. Because of this late reporting, the notification could not be analysed in detail. In addition, Eurostat had significant doubts about the quality of the figures provided.
The October reporting by the Greek authorities is exceptional in terms of procedure, but it is neither without precedents nor an isolated episode:
- Following evidence of widespread misreporting of deficit and debt data by the Greek authorities during the previous years, in November 2004 a Report on the revision of Greek government deficit and debt figures was released by Eurostat, which showed that figures had been misreported in the years preceding 2004 involving no less than 11 separate issues.
- On five occasions between 2005 and 2009 Eurostat expressed reservations on the Greek data in the biannual press release on deficit and debt data.
- Over the last eight years, whenever the Greek EDP data have been published without reservations, this was very often the result of Eurostat interventions before or during the notification period in order to correct mistakes or inappropriate recording.
In its conclusions of 10 November 2009, the Economic and Financial Affairs (ECOFIN) Council expressed regret2 for the renewed problems with the figures reported by the Greek Government and called on it to restore confidence in Greek statistical information and the related institutional setting. Furthermore, the ECOFIN Council also invited the Commission to produce a report and propose the appropriate measures to be taken in this situation.
As the quality problems of Greek statistical information are not only linked to public finance data - National Accounts (GDP) figures have also been significantly revised in the past - and as the events of October 2009 cannot be deemed as exceptional, the report will not deal only with statistical methodological issues. Rather, an analysis of the governance and institutional framework of the Greek system for the production of statistical information is also necessary.
The report is organized as follows: section 2 provides information on the institutional framework of Greek statistics. Section 3 provides a chronology of the main events concerning Greek EDP statistics from 2004 to 2009. Section 4 summarizes the methodological issues which led to revisions of data between 2005 and 2009 and in particular between April 2009 and October 2009. Finally, section 5 provides the conclusions of this report.
2Institutional framework
2.1 European Statistical System and quality assurance
Official Statistics in Europe are organised in the 'European Statistical System (ESS)', a partnership between the statistical authority of the European Union, which is the Commission (Eurostat), and the national statistical institutes (NSIs) and other national authorities responsible in each Member State for the development, production and dissemination of European statistics. Member States collect data and compile statistics for national and EU purposes. The ESS functions as a network in which Eurostat’s role is to lead the way in the harmonization of statistics in close cooperation with the national statistical authorities.
Quality assurance in the ESS has emerged from developments over many years in the field of statistics. There is a balance to be found between the objectives of supervision and control by Eurostat and the constraints arising from the principle of subsidiarity and the autonomy of Member States in the way they organise and produce Official Statistics. Although the responsibility for monitoring statistical data is vested in the Commission (and in the case of the Excessive Deficit Procedure the Commission 'provides' the data), the Commission does not directly compile statistics or control their production in the Member States. In this respect, it depends largely on the data compiled and reported by the Member States, as well as the administrative ability, good will and loyal co-operation of the respective national authorities. Appropriate institutional frameworks ("governance"), respect of principles, compliance with production methods, and plausibility checks for transmitted data are the available tools for quality assurance in European statistics.
- Member States are bound to cooperate in good faith and in conformity with the statistical principles set out in Regulation 223/20093. These principles require that statistics are produced in an independent manner, free from any pressures from political or interest groups, and also in a systematic, reliable and unbiased manner using professional and ethical standards. Scientific criteria must be used for the selection of sources, methods and procedures. As a result, it is a primary obligation on the Member States to provide statistical data in accordance with such principles and minimum standards for the independence, integrity and accountability of national statistical authorities and to take all necessary measures to ensure that their institutional and organizational set-up respects the rule of law.
- Member States must organise the governance in the National Statistical System and specify the corresponding roles of the institutions involved in a precise manner, determining which authority is in charge of coordination, which institutions belong to the group of producers and which institutions are outside the remit of Official Statistics.
- Methodological prescriptions and definitions (e.g. in the European System of National Accounts) have to be applied by the Statistical Authorities.
- Eurostat is in charge of supervising compliance with European law (application of methods and delivery of statistical data) and of assuring the plausibility of statistical information.
This quality assurance system and its features is however based on assumptions which have shown to be subject to shortcomings and limitations, of particular relevance in the case of Greece:
- The partners in the ESS are supposed to cooperate in good faith. Deliberate misreporting or fraud is not foreseen in the regulation.
- The governance structure has to be implemented in accordance with the administrative environment of a Member State; there is no single model for the purpose of ensuring 'good' governance.
- Compliance with the methodological procedures of European Statistics is only binding for institutions which are considered as Statistical Producers (defined in Regulation 223/09); insufficient accounting and bookkeeping practices outside the statistical area are not easily detectable by statistical control.
- The correctness of data delivered from different national sources and respondents is also outside Eurostat's power of control. Hence, powers granted to Eurostat by Regulation 223/2009 with respect to the governance structure and to control the quality of data outside the narrow limits of 'statistical producers' are almost nil.
2.2European Statistics Code of Practice
After the Greek misreporting of the EDP data in 2004, the ECOFIN requested the establishment of minimum standards for independence, integrity and accountability of national statistical authorities; however in 2005 the focus was shifted to the independence of Eurostat. A compromise was reached in November 2005, when the ECOFIN concluded that an external body should be established to focus its monitoring of compliance with these standards by Eurostat and the ESS as a whole.
The Commission adopted a global strategy for strengthening the European Union governance of fiscal statistics along three lines of action: building up the legislative framework; improving the operational capacities of the relevant Commission services, and establishing the 'European Statistics Code of Practice'4. The Code has a self-regulatory character.
In 2007 the Commission adopted a proposal for regulation regarding the establishment of an external body to monitor compliance with the Code by national statistical authorities and Eurostat. The European Statistical Governance Advisory Board (ESGAB) was established by a European Parliament and Council Decision in 2008 to provide an independent overview of the ESS as regards the implementation of the Code. ESGAB started to work in March 2009 and published its first report in November 2009, which includes some general recommendations regarding the institutional set up in reaction to the first information about the new Greek case: 'An appropriate institutional framework is crucial in order to safeguard the professional independence of statistical authorities. Suspicions of interventions affecting the data produced need to be further investigated. Moreover, the procedures for the appointment and dismissal of Heads of National Statistical Institutes (NSIs) should be transparent and kept separate from political mandates.' The report also stresses that 'a stronger commitment from top management in the statistical offices and a stronger adherence to common quality standards at the level of the ESS will be of essence'.
2.3Institutional and legal aspects of EDP reporting
Article 126 of the Treaty on the Functioning of the European Union (TFEU) specifies that Member States shall avoid excessive government deficits. Under Protocol (No 12) on the excessive deficit procedure, the Member States regularly report to the Commission their actual and planned deficit and debt levels; the data are 'provided' by the Commission. Pursuant to Article 6(2) of Regulation (EC) No 223/2009, the Commission (Eurostat) shall ensure the production of European statistics according to established rules and statistical principles. In this respect, it shall have the responsibility for deciding on processes, statistical methods, standards and procedures. Under Regulation (EC) No 479/2009, the Commission is also entrusted a task to ensure compliance with the required budgetary discipline by monitoring the development of the budgetary situation and of the stock of government debt in the Member States. Coherence and comparability of data in the fiscal statistics framework rely on the correct implementation by the Member States of the legal acts relevant in the context of the EDP, such as Council Regulation (EC) No 2223/965 (ESA 95) and Council Regulation (EC) No 479/20096 establishing the statistical reference system for standards, definitions and accountancy provisions.
Article 8(1) of Regulation (EC) No 479/2009 sets forth that the Commission (Eurostat) shall regularly assess the quality both of actual data reported by Member States and of the underlying government sector accounts compiled according to ESA 95 (hereinafter referred to as government accounts). Quality of actual data means compliance with accounting rules, completeness, reliability, timeliness, and consistency of the statistical data. Further, according to Article 8(2) of Regulation (EC) No 479/2009, Member States shall provide the Commission (Eurostat), as promptly as possible, with the relevant statistical information requested for the needs of the data quality assessment, without prejudice to the provisions relating to statistical confidentiality.
Following the Greek case in 2004, and a request by the Council to 'strengthen the monitoring of the quality of the reported fiscal data', the Commission proposed amendments to the existing Council Regulation for the quality of EDP data (Regulation 3605/93).7 Council Regulation 2103/20058 granted Eurostat strengthened control powers, though more limited than initially requested by the Commission. These powers work within strict constraints, notably:
- Methodological visits are exceptional ("should only be undertaken in cases where the Commission (Eurostat) identifies substantial risks or potential problems with the quality of the data, especially where it relates to the methods, concepts and classifications applied to the data, which Member States are obliged to report"). In contrast, the Commission proposal foresaw such visits in a systematic way to all Member States.
- Even more importantly, also in contrast with the Commission proposal, there is no general reference to the obligation of Member States to provide Eurostat with access to 'the information' requested for the purposes of the data quality assessment.
- Crucially, Eurostat is restricted to statistical matters in its work; the institutional setting is out of their scope of action 'The methodological visits should not go beyond the purely statistical domain', and that the interlocutors of Eurostat should be 'the services responsible for the excessive deficit procedure reporting'. This provides Member States with potential arguments for restricting the access to information.
Member States are nevertheless responsible for taking 'all necessary measures to facilitate the methodological visits' and 'making documents available to justify the reported actual deficit and debt data and the underlying government accounts.' This includes 'national authorities which have a functional responsibility for the control of the public accounts' (i.e. national audit offices). Thus, within the constraints identified above (and allowing for some potential contradiction between different parts of the Regulation), Eurostat in principle has the legal means to secure access to the necessary documents and at least to the auditors of public accounts. The question remains, however, whether it has the necessary means to enforce the respect of these competences.
In the specific case of Greece, Eurostat has made a quasi-permanent and extensive use of the existing powers to monitor Greek EDP data, devoting far more resources and more missions to the country than to any other EU Member State. It is the only Member State that has received methodological visits. Following these visits, extensive action plans have been put in place based on an analysis of all information made available by the Greek authorities. However even this activity was unable to fully detect the degree of interference in the Greek EDP data.
All in all, although Council Regulation 2103/2005 fell short of what was proposed by the Commission, enlarged powers would only have contributed to reducing the risk that a country could report incorrect data, not to eliminating it.
2.4 Governance and institutional framework for EDP reporting in Greece
The institutions in Greece responsible in the reporting for each table in the EDP notification, are those reported by the Greek authorities in the April 2009 EDP notification: the National Statistical Service of Greece (NSSG), the Ministry of Finance (MOF, through the General Accounting Office, GAO), the Single Payment Authority (which is also part of the Ministry of Finance) and the Bank of Greece (BOG). Specifically, the NSSG is responsible, together with the MOF, for the deficit reporting, while the MOF is fully responsible for the debt figures.
The Ministry of Finance is responsible for the compilation of the tables for central government, while the National Statistical Service of Greece is responsible for the data concerning local government and the social security sector. The SPA is responsible for providing data on EU grants. The Bank of Greece is responsible for the compilation of a full set of financial accounts and financial balance sheets for the whole economy9. This means that it is also responsible for the correctness of the EDP data, notably for the compilation of financial transactions and balance sheets for the general government sector. This division of responsibilities is not uncommon in the EU.
The overall governance structure and institutional framework has nevertheless remained unclear.10 The government has nominated only the NSSG and the BOG as Statistical Authorities working under the umbrella of the European Statistical Law11, but not the MOF (i.e. General Accounting Office) or the SPA (Single payment Authority), which are however major actors in the compilation of EDP data. The revision of the statistical law was repeatedly postponed. Moreover, contrary to the commitment of the Greek statistical authorities to Eurostat in September 2008 (see section 3.3), a statistical council has not been set up yet.
During the EDP methodological mission of November 2009, as a direct consequence of the reservation introduced on Greek EDP data, the shortcomings of the institutional setting were again addressed. All participants at the meeting, including the Greek authorities (NSSG and the GAO), concluded that the present institutional setting could not guarantee that EDP notification figures would be fully free from political interference. Moreover, a worrying lack of accountability and unclear responsibilities of the entities involved in compiling EDP statistics was identified, a problem of potential political/external interference with the EDP figures to be transmitted to the Commission.
To address this issue:
- The responsibilities of the different entities involved need to be clarified. Co-operation between agencies needs to be formalised, possibly through a memorandum of understanding, and it must be clarified which agency is responsible for which data in EDP notifications.
- Responsibilities need to be personalised. Senior officials should be responsible for the data produced under their responsibility.
- In this context, figures should only be reported if the agency providing the raw data for compiling the cell in EDP assures the reliability of that figure though a process of certification and signature.
- The provisions included in the Code of Practice must be respected.
Furthermore, making the NSSG independent, through the revision of the current law on statistics, will be a necessary, but not a sufficient condition to ensure the principles above.
In addition to these aspects and with reference to the findings of the Peer Review in 2007 co-operation and professional knowledge transfer between Greek statistics and the partners in the European Statistical System should be intensified.
3 Chronology of main events
The reliability of Greek government deficit and debt statistics has been the subject to continuous attention for several years. The Greek general government data reported by the authorities have been persistently contested by Eurostat, far more frequently than for any other Member State. As a result of the observed poor quality of the Greek general government data, reservations on the reported debt and deficit figures were repeatedly expressed in the bi-annual Press Releases (see Annex 3 for the list and text of reservations since March 2005).
The following subsections provide an account of the revisions of the data and actions undertaken to help the Greek authorities improve the quality of government statistics and to comply with the existing rules and guidelines.
3.1The 2004 methodological report
3.1.1Exceptional revisions in the September 2004 EDP notifications
In March 2004 Eurostat expressed a reservation on the quality of the Greek figures12. In September 200413 the Greek general government deficit and debt data were significantly revised for the whole reporting period 2000-2003. In the subsequent press release on government deficit and debt in the EU and in the euro-zone, Eurostat explained in detail why it had considerably revised the data for the previous years (from 1997 onwards)14. In particular, the revision of the data notified for the year 2003 between the March 2004 and September 2004 notifications reached almost three percentage points of GDP for the deficit and more than seven percentage points of GDP for the debt. The deficits notified for the years 2000, 2001 and 2002 were also revised upwards by more than two percentage points of GDP. These significant upward revisions were the result of earlier actions undertaken by Eurostat as well as of the initiative of the Greek government in spring 2004 to launch a thorough fiscal audit.
Table 1 shows the extent of the revision of the Greek figures from March to September 2004, split between the main components of the revision.
Table 1. Main components of the revision of Greek data between March and September 2004
| 2000 | 2001 | 2002 | 2003 | |
| DEFICIT | % GDP | % GDP | % GDP | % of GDP |
| March 2004 | -2.0 | -1.4 | -1.4 | -1.7 |
| Tax revenue | 0.9 | |||
| Payments from the EU | 0.3 | |||
| Reclassification of payments from the Postal Bank | 0.2 | |||
| Military expenditure | 1.9 | 1.2 | 1.7 | 0.7 |
| Surplus of Social Security Funds | 0.0 | 1.0 | 0.4 | 0.6 |
| Under recording of interest | 0.3 | 0.1 | 0.1 | 0.1 |
| September2004 | -4.1 | -3.7 | -3.7 | -4.6 |
| DEBT | ||||
| March 2004 | 106.1 | 106.6 | 104.6 | 102.6 |
| Capitalised Interest | 4.5 | 4.2 | 3.9 | 3.4 |
| Consolidating Assets of Social Security | 3.2 | 3.8 | 3.8 | 3.7 |
| 0.1 | 0.1 | 0.2 | 0.1 | |
| September 2004 | 114.0 | 114.7 | 112.5 | 109.9 |
3.1.2The 2004 methodological report and the infringement procedure
On 22 November 2004, Eurostat published a comprehensive report on the revision of the Greek government deficit and debt figures. The report showed that, in the years preceding 2004, the Greek authorities had misreported deficit figures to the Commission by not following ESA 95 rules. The misreporting involved no less than eleven separate issues.
In this context, an infringement procedure was opened by the Commission in 2004 against Greece concerning EDP reporting, due to methodological problems in the recording of taxes and social contributions, the surplus of social security funds and hospitals and the recording of debt assumptions and military expenditures. The procedure was closed only in 2007 (see Box 2 and Annex 6 of this report).
In 2005 Eurostat conducted three dialogue visits in Greece in order to verify the compliance of the Greek deficit and debt data with the existing methodological requirements. These dealt mainly with hospital liabilities, debt assumptions, EU grants and swaps, alongside many other issues.
3.2Greek reporting between the EDP methodological visit of 29 May-2 June 2006 and 27-29 September 2006, the action plan and follow-up actions (up to 17 July 2008)
Following the notification by the Greek authorities in spring 2006 (see Table 2 for the successive revisions of notified data), in April 2006 Eurostat again expressed a reservation on the Greek government finance statistics data. This was followed by two methodological visits in May and September 2006 and by a methodological report prepared by Eurostat, agreed with the Greek authorities and published on the Eurostat website. The report contains a detailed list of short, medium and long term recommendations, aimed at ensuring a better quality and reliability of data on the balances of social security funds, local government and extra-budgetary funds, as well as the recording of the transactions with the EU budget according to the existing rules.
Table 2 – Government deficit and debt data notified by Greece between 2005 and 2009
| GREECE Deficit (% of GDP) | 2001 | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | |
| Notification of | |||||||||
| 2005 | |||||||||
| March * | -3.6 | -4.1 | -5.2 | -6.1 | - | - | - | - | |
| September * | -6.1 | -4.9 | -5.7 | -6.6 | - | - | - | - | |
| 2006 | |||||||||
| April | - | -4.9 | -5.8 | -6.9 | -4.5 | - | - | - | |
| October | - | -5.2 | -6.1 | -7.8 | -5.2 | - | - | - | |
| 2007 | |||||||||
| April | - | - | -6.2 | -7.9 | -5.5 | -2.6 | - | - | |
| October | - | - | -5.6 | -7.3 | -5.1 | -2.5 | - | - | |
| 2008 | |||||||||
| April | - | - | - | -7.4 | -5.1 | -2.6 | -2.8 | - | |
| October | - | - | - | -7.5 | -5.1 | -2.8 | -3.5 | - | |
| 2009 | |||||||||
| April | - | - | - | - | -5.1 | -2.8 | -3.6 | -5.0 | |
| October | - | - | - | - | -5.2 | -2.9 | -3.7 | -7.7 |
| GREECE Debt (% of GDP) | 2001 | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | |
| Notification of | |||||||||
| 2005 | |||||||||
| March * | 114.8 | 112.2 | 109.3 | 110.5 | - | - | - | - | |
| September * | 114.4 | 111.6 | 108.8 | 109.3 | - | - | - | - | |
| 2006 | |||||||||
| April | - | 110.7 | 107.8 | 108.5 | 107.5 | - | - | - | |
| October | - | 110.7 | 107.8 | 108.5 | 107.5 | - | - | - | |
| 2007 | |||||||||
| April | - | - | 107.8 | 108.5 | 107.5 | 104.6 | - | - | |
| October | - | - | 97.9 | 98.6 | 98.0 | 95.3 | - | - | |
| 2008 | |||||||||
| April | - | - | - | 98.6 | 98.0 | 95.3 | 94.5 | - | |
| October | - | - | - | 98.6 | 98.8 | 95.9 | 94.8 | - | |
| 2009 | |||||||||
| April | - | - | - | - | 98.8 | 95.9 | 94.8 | 97.6 | |
| October | - | - | - | - | 100.0 | 97.1 | 95.6 | 99.2 |
* Regulation 3605/93 requested transmission of data before 1 March and 1 September.
In this context, agreement was reached with the Greek authorities on an action plan for the implementation of the Eurostat recommendations (Box 1 and Annex 4). It was also agreed that the effective monitoring of the implementation of the action plan would require the regular submission of a formal progress report by the Greek authorities to Eurostat.
Box 1. The Action Plan
As a follow up on the reservation expressed on Greece's fiscal data by Eurostat on 24 April 2006, which highlighted the persistence of issues of a structural and systemic nature affecting the Greek government accounts1, an EDP methodological visit took place to Athens from 29 May to 2 June 2006 and from 27 to 29 September 2006. The findings of the June 2006 methodological visit to Greece were the basis for an action plan, agreed in detail between Eurostat and the Greek authorities.
The initial action plan involved actions to address the provision of appropriate data on social security funds, local government and extra-budgetary funds, and on transactions with the EU budget. was subsequently enlarged during the methodological visits undertaken in 2008 to cover the appropriate corrections to be imputed with respect to EU flows and to solve the problem of the discrepancy existing in central government for 2006 and 2007: this implied reviewing in detail the compilation of budgetary central government statistics, and investigating a more complete use of the Survey for the compilation of extra budgetary funds statistics (see Annex 4).
Eurostat has closely monitored the implementation of the action plan.
At the time of the EDP notification of April 2007, some of the actions had started to be implemented and some partial results were obtained, mainly relating to consistency checks for the largest units inside the Survey for extra-budgetary central government, local government and social security funds, improving the coverage of the Survey, amending the transitional items in the table showing the conversion of the Survey results into national accounts concepts, changing the calculation of receivables relating to transactions with the EU budget.
Further exchanges with the Greek authorities in summer 2007 raised Eurostat concerns about the implementation of specific actions of the action plan, in particular on the recording of taxes and social contributions and on the income and expenditure of social security funds of hospitals.
Exchanges and missions up until the methodological visit undertaken by Eurostat to Athens on 16-18 November 2009 had led to an updated list of actions which was provided by the Greek authorities. Although a number of actions which should have been already implemented are still in progress or in an incipient phase (in particular in the area of improving timeliness and coverage of the survey of social security funds and hospitals, verifying internal consistency of each questionnaire and requiring the correction and resending of systematically unbalanced questionnaires) the majority of the actions to be undertaken by the Greek authorities in the context of the action plan have been implemented. Some of the remaining actions are outside the scope of the NSSG and will require political willingness from Government, specifically the implementation of existing legal acts and reinforcement of existing legislation for social security funds and local authorities, and the introduction of balance-sheet reporting and accrual based accounting in public bodies. Nevertheless, even full compliance by the Greek authorities with the action plan (which foresaw actions to be taken only at a technical-statistical level), would not have prevented the deliberate misreporting of figures by the Greek authorities in 2009.
1 The reservation also referred to Eurostat Press Release 120/2005 of 26 September 2005, which included the following reservation on the Greek data: 'Some pending issues remain for the recording of EU transactions, the accounts of social security and the amounts of other receivables and payables for the years 2002-2004'.
In May 2007, the infringement procedure was closed (Box 2).
Box 2. The closure of the infringement procedure in 2007
The infringement procedure, which had been opened in 2004, was closed in 2007 following the conclusion that in view of the measures undertaken by the Greek authorities the specific grounds for the infringement procedure no longer existed (see Annex 6). In addition, the following general considerations were taken into account by the Commission when adopting the decision:
- The letter of formal notice indicated that there were persistent systemic problems in the relevant departments of the Greek administration.
- Concrete measures had been taken by the Greek authorities in order to try solving these issues. Their implementation was to be followed by Eurostat.
- Despite the recent improvement in the statistical processes and good co-operation between Eurostat and the national statistical authorities of Greece, issues of a structural and systemic nature related to the Greek government accounts remained.
- Eurostat noted that improvements had been made in the compilation of government finance statistics. Eurostat would finalize and publish its report including an action plan with a timetable for the implementation of the remaining issues.
The fact that the strict conditions on the basis of which the infringement procedure had been launched no longer applied/existed, did not therefore signify that Greek data were compliant with all the applicable rules and procedures, as far as their quality was concerned. It only implied that that particular legal instrument could no longer be applied, which raises some questions about the effectiveness of the instrument in addressing complex issues combining methodological and institutional aspects and points to the need to take account of this complexity in the future. In fact, between 2005 and 2009, in over 10 EDP notifications, Eurostat introduced reservations on the quality of the data submitted by the Greek authorities no less than five times, far more than for any other Member State. In the specific case of Greece, Eurostat has had a quasi-permanent and extensive use of the existing powers to monitor Greek EDP data. Greece is the only Member State which has received methodological visits so far. Following these visits, extensive action plans have been put in place based on the analysis of all information made available by the Greek authorities. However even this activity was unable to detect the level of (hidden) interference in the Greek EDP data.
In particular, after the closure of the infringement procedure at the end of 2007, Eurostat issued a reservation on the quality of the Greek data in the April 2008 notification and validated the notifications of October 2008 and April 2009 only after it intervened before and during the notification period to correct mistakes or inappropriate recording, with the result of increasing the notified deficit in both instances. As an example, Eurostat's methodological missions in 2008 resulted in an increase of the 2007 deficit figure notified by the Greek authorities, from 2.8% to 3.5% of GDP. The list of reservations introduced by Eurostat between 2005 and 2009, and the text of those reservations, can be found in Annex 3.
According to the agreed action plan and commitments of the Greek authorities, Eurostat requested, in a letter of 27 June 2007, the Greek authorities to provide a report on the progress made with respect to the recommendations listed in the report of the EDP methodological visit. The answer of the Greek authorities (3 July 2007) outlined the main steps taken by the Greek authorities for the implementation of the action plan, but only in respect of the long-term recommendations. Therefore in a letter of 13 July 2007 a more detailed overview of the situation regarding the short and medium-term actions was requested. In addition, the Greek authorities were also asked to provide detailed information on timeliness-coverage of their annual Survey by subsector.
In reaction to the further response of the Greek authorities (31 July 2007), in a letter of 7 September 2007 Eurostat expressed the view that despite the progress reported by the Greek authorities, some key recommendations mainly relating to internal consistency checks (for the Survey) remained pending and it invited the Greek authorities to send an exhaustive progress report by the end of January 2008. Furthermore it requested the Greek authorities to attach a report on the issue of internal consistency of the October 2007 EDP notification.
After this exchange of letters, some information on the progress made with respect to the timeliness, coverage and internal consistency of the Survey was provided during the October 2007 EDP notifications and a more exhaustive progress report was finally sent on 17 July 2008.
3.3Greek reporting in 2008: the follow-up EDP methodological visit of 2-3 June 2008 and second EDP methodological visit of 15-19 September 2008
Following the notification of the Greek authorities in April 2008, Eurostat once again expressed reservations on the notified Greek deficit, in particular with respect to issues relating to the recording of EU grants in 2006 and 2007, the existence of a substantial statistical discrepancy in 2007 of 0.6% of GDP and the insufficient coverage of source data for extra-budgetary funds, local government and social security funds achieved for the first estimate of the 2007 balance.
This led to another methodological visit to Greece on 2-3 June 2008, the main aim of which was to clarify the issues on which a reservation had been issued. The visit led to a published report ("Main conclusions and action points")15 sent by Eurostat to the Greek authorities on 8 July 2008 containing the agreed recommendations and implementation actions. Such actions were integrated in the action plan.
On 11 September 2008, in the follow-up of the 2007 Peer Report, a Eurostat high-level mission discussed actions aimed at further improving the quality of data provided in the EDP framework, as well as at implementing the improvements in the institutional setting of the Greek Statistical System as foreseen in the recommendations of the 2007 Peer Review. The Minister of Finance and the Head of the NSSG explicitly committed to consolidate the institutional arrangements in Greece in line with recommendations made in the Code of Practice Peer Review of Greece. In particular, the Minister of Finance and Economy and Secretary General of NSSG stressed their commitment to fulfil their obligation to provide reliable EDP figures, and to co-operate without restrictions and in full transparency with the objective to fully address the remaining open questions. In particular, the revision of the Greek statistical law and the constitution of a user council were said to be in a late stage of realisation. Eurostat underlined that full compliance with the Code of Practice is a precondition for the trust in Greek statistics and consequently in the statistics of the ESS.
On 15-19 September 2008 a new methodological visit to Greece took place. The aim of the visit was more specifically to agree on the appropriate corrections with respect to EU flows which had not come to a satisfactory conclusion in the previous methodological visit and to solve the problem of the sizeable statistical discrepancy existing in central government data for 2006 and 2007. The agreed findings and short and medium term actions are included in the published Eurostat Report on this methodological visit to Greece16. The discussions led to a considerable increase in the deficit estimate for 2007 notified by Greece (from 2.8% notified in April to 3.5% of GDP) and consequently to the lifting of the reservations on Greek EDP data that had been made in April.
3.4Greek government deficit and debt figures: revisions April 2009 – October 2009
In April 2009 Eurostat validated the data notified by the Greek authorities, which reported a deficit for 2008 of 5% of GDP (the figure initially notified by Greece was 4.8% of GDP). With hindsight, this decision does not appear to have been appropriate; it was based, however, on the knowledge available at the time and took place after the Greek authorities, under advice from Eurostat, which had raised doubts on the surplus of the social security sector, reduced by half the positive statistical discrepancy of social security, imputing the difference to a reduction of the surplus of the sector (raising the deficit of general government by slightly more than 0.2% of GDP).
After the April notification and before the press release was issued, Eurostat contacted the Greek authorities to clarify an issue relating to the recording of debts of public hospitals to their suppliers. In response, the Greek authorities sent official documents of the Ministry of Health certifying that the amounts reported in the notification were those recognised by the Greek government (Annex 5).
In the context of the April 2009 EDP notification, Eurostat again drew the attention of the Greek authorities to the actions agreed in the context of the action plan. Although in their reply the Greek authorities pointed to elements of improvement (notably regarding the survey coverage of the social security funds), many issues remained unresolved (such as the classification of extra budgetary accounts and the classification of government units).
On 27 April 2009, the Council recommended ("Council Recommendation of 27 April 2009 to Greece with a view to bringing an end to the situation of an excessive government deficit, in accordance with Article 104(7) of the Treaty for Greece") that: 'The Greek authorities should continue efforts to improve the collection and processing of statistical data and in particular general government data, notably by enhancing the mechanisms that ensure the prompt and correct supply of the general government data required by the existing legal framework'.
When the Greek authorities submitted 2 October 2009 notification, which presented a large revision compared to the April notification, Eurostat asked the Greek authorities to clarify the observed discrepancy with the quarterly financial accounts of general government. It also asked for additional information on the revision to accrued EU revenue in 2008, missing details in the 'other accounts payable' in 2007 and 2008, and noted that the classification of the extra budgetary accounts had not yet been completed. 17
The Greek authorities did not reply to the second request for clarification and asked for an extension of the deadline set by Eurostat on 13 October 2009. An incomplete answer was finally received on 21 October, and additional information on 22 October, after the publication of EDP data by Eurostat.
Particular characteristics of the October 2009 notification were:
- A revised EDP notification was received by Eurostat on 21 October, the afternoon before the publication of its press release;
- Questions from Eurostat went unanswered for no less than 9 days, an abnormally long period, between 12 and 21 October;
- In the period between 16 and 21 October, the General Secretary of the NSSG repeatedly contacted Eurostat claiming political interference over the provision of figures and with the sending of a revised notification18.
More specifically, with regard to compilation issues, the investigation of Eurostat has revealed weaknesses in the national services principally responsible for EDP data: the National Statistical Service of Greece (NSSG), the General Accounting Office (GAO) and the Ministry of Finance (MOF). In the context of the April, 2 October 2 and 21 October 2009 EDP notifications transmitted by the Greek authorities, there is evidence that the NSSG modified the figures for the correction on EU grants communicated by the Single Payment Authority and modified the results of the survey on revenues/expenditures of hospitals, decreasing the amount of outstanding liabilities resulting from the survey and reducing government deficit in 2007 and 2008. The GAO (itself an entity part of the Ministry of Finance) transmitted to the NSSG wrong data on swaps cancellation, on extra-budgetary accounts and on accrued interest as an expenditure of government, reducing government deficit in April and on 2 October 2009. The MOF took in October 2009 decisions on the recording of transactions which had taken place in 2008 involving government, changing the recording of operations ex-post even if the transactions had resulted from decisions taken in 2008. Moreover the MOF instructed the NSSG to record in the year 2008 a liability for hospitals equal to € 2.5 bn in the EDP notification of 21 October 2009, although no evidence and no justification for the time of the recording of the transaction was provided.
The revisions introduced from the April 2009 to 2 October 2009 and 21 October 2009 EDP notifications are shown in Table 3.
4 Summary of methodological issues
This chapter summarises the main methodological issues discussed in this report, pointing out whether such issues had also been discussed in the past (in particular in the Report by Eurostat on the revision of the Greek government deficit and debt figures, hence forth the “2004 Eurostat report”) and providing information on the provisional conclusions reached on each issue. Additional information for each of the issues detailed in each chapter is provided in Annex 1 of this report.
As far as a possible parallel between the 2004 and 2009 situations is concerned, as can be deducted from the following paragraphs, there are some common methodological features between the 2004 and 2009 episodes. In both cases, in the aftermath of political elections, substantial revisions took place revealing a practice of widespread misreporting, in an environment in which checks and balances appear absent, information opaque and distorted, and institutions weak and poorly coordinated. The frequent missions conducted by Eurostat in the interval between these episodes, the high number of methodological visits, the numerous reservations to the notifications of the Greek authorities, on top of the non-compliance with Eurostat recommendations despite assurances to the contrary, provide additional evidence that the problems are only partly of a methodological nature and would largely lie beyond the statistical sphere.
The five reservations expressed between 2005 and 2008 on the quality of Greek EDP data, were motivated by the inappropriate recording of EU grants (twice), the existence of a non-negligible statistical discrepancy, the accounts of social security, doubts on the recording of the amounts of other receivables and payables and the insufficient coverage of source data for extra-budgetary funds, local government and social security funds achieved for the first estimate in the context of the EDP notification of April. All these issues were subsequently clarified between Eurostat and the Greek authorities and closed accordingly. However, some of these issues (the recording of EU grants, the accounts of social security and the liabilities of hospitals) resurfaced in 2009:
Nevertheless, in the case of EU grants, there is a significant difference between the situation in 2005–2008 and the one in 2009. Between 2005 and 2008, the revision of data undertaken by Eurostat was due mainly to methodological reasons, while on 2 October 2009 there seems to have been no methodological issue at stake but simply misreporting of the data communicated by the Greek authorities. In the case of social security, the revisions in 2005–2008 were also due mainly to methodological issues. On the contrary it is not yet clear, whether the inflated surplus of social security reported in the April 2009 notification was only the result of methodological uncertainties.
In the case of hospitals liabilities, the issue in 2005 referred to past expenditures of hospitals that had never been recorded. Now, it appears that around € 1 bn of hospitals liabilities, reported in the survey on hospitals, were ignored by the Greek authorities in the April and 2 October 2009 notifications, although the misreporting had actually started, for smaller amounts, already in 2007.
4.1Working balance of the state budget
The issue of the recording of inflows and outflows of the State budget and their link with the working balance was not included in the 2004 Eurostat report, although it was intensively discussed during the 2006 and 2008 EDP methodological missions.
In the EDP notification dated 21 October 2009, the working balance of the government budget was increased by € 722 mn for the year 2008 by the GAO (from around – €13.9 to – 14.6 bn). According to the explanations received by the Greek authorities during the EDP methodological visit of Eurostat held on 16–18 November 2009 (henceforth the November 2009 methodological visit), the revision was due almost in its entirety to the inclusion of a previously neglected expenditure (for an amount of € 710 mn) equal to the transfer from the State budget to the social security fund of a state-owned company, OEP-DEH.
Following the subsequent submission of information requested by Eurostat to the Greek authorities, it is still not clear (the NSSG and the GAO having provided mutually contradictory information on the issue):
- whether privatisation proceeds have or have not been included in the working balances of the previous years;
- whether the transfer to OEP-DEH has been recorded as a government expenditure or not (whether it was the object of consolidation of intra-government flows);
- whether other intra-government flows have been properly consolidated in the past.
Notwithstanding the above, Eurostat currently has major doubts on this revision undertaken by the GAO in the 21 October notification.
4.2Revision of Treasury Accounts
In the EDP notification of 21 October 2009, the amount of other accounts payable of the Greek central government increased by €192 mn in 2008 and by minor amounts for the years 2005 to 2007, increasing government deficit for all years. This resulted from the revision of the Treasury Accounts which exist in Greece to support specific operations which do not transit via the budget, such as for military expenditures or payments on guarantees called.
Although the issue was never discussed in detail before 2004 and was not included in the 2004 Eurostat report, the recording of flows in Treasury Accounts was discussed in detail during the following years and particularly in 2008. These accounts have been a source of statistical discrepancy in the past, due to the fact that many sub-accounts of the Treasury Single Account have been excluded from the EDP notification. Eurostat requested in 2008 that the Greek statistical authorities decide on the final classification of such accounts (determining whether they were owned by government units or not). Although on 10 October 2009 the GAO informed Eurostat that it would not be possible to conclude the task of investigating and identifying each account before the end of the year, by 16 October 2009 the task had been already completed, apparently due to the fact that the new political authorities had instructed the GAO to complete the examination of the Treasury Accounts as soon as possible.
Given the above, Eurostat is not in a position at present to assess whether each individual extra-budgetary account has been correctly classified and whether the correction undertaken for government accounts is correct. Although the issue seems to be of a methodological nature, it is surprising that in spite of the information provided previously by the GAO in October about the impossibility to conclude the work before the end of the year, it was seemingly possible to undertake it in only a few days after the change of government. In this context, the modalities and the precise amount of the revision will have to be ascertained by Eurostat.
4.3 Revenues from abolished extra-budgetary accounts
This issue was not among those included in the 2004 report.
In the EDP notification of 21 October 2009, the government deficit of 2008 increased by € 300 mn due to the exclusion from revenue of written-off extra-budgetary accounts. These are accounts owned by government which are not included in the working balance. The origin of the correction was that, in 2008, the government decided to abolish some of these accounts and to transfer the outstanding amounts of funds into the State budget. Such amounts were included by the GAO incorrectly, in the April and 2 October EDP notifications, as revenue in the working balance of table 2A. Since this flow involved only government entities, it should have been consolidated inside government and the amount should either not have been included in the working balance of central government or, if included, a corresponding adjustment with the opposite sign should have been imputed in table 2A, which was not the case.
Although the exclusion of these amounts from government revenue in the 21 October EDP notification seems correct, it remains to be checked whether a corresponding outflow was recorded (in order to neutralise the flow) or not recorded, in the adjustment item reflecting the balance of extra-budgetary accounts. Moreover, it remains to be seen how and whether flows resulting from similar operations have been recorded in the past.
This is to be considered as a case of deliberate misreporting of figures by the GAO in the April and 2 October 2009 EDP notifications.
4.4Swaps write-offs
The issue of swaps write-offs was not among those included in the 2004 report. During the years 2005 to 2008, on the contrary, the issue of the cancellation of swaps and of the treatment of their remaining stream of interest was discussed between Eurostat and the Greek authorities.
In the 21 October 2009 EDP notification, figures were revised by € 210 mn for the year 2008, increasing the government deficit, due to a swap cancellation. This, however, was simply an issue of misreporting of figures by the GAO in the April and 2 October 2009 EDP notifications, rather than a methodological issue or the introduction of a mistake.
In short, in the context of swaps cancellations, lump sums received by government should not be considered in national accounts as government revenue. As such amounts were included in the working balance of the government, a negative adjustment should have been included in the EDP tables for the same amount. There is evidence that the GAO knew already in April 2009 both the amount of the correction to impute and how it should have been neutralised, but did so only on 21 October, underestimating as a consequence government deficit in the April and in the 2 October EDP notifications.
This is to be considered as a case of deliberate misreporting of figures by the GAO in the April and 2 October 2009 EDP notifications.
4.5Adjustment for interest payments
The recording of interest was one of the issues included in the Eurostat 2004 report. However, at the time the issue was a methodological one and dealt with the recording of capitalised interest in those cases in which the Greek government had been issuing bonds which did not carry any interest during the first years.
On the contrary, the issue in 2009 related simply to the fact that figures on accrued interest which constituted government expenditure, although correctly calculated since March 2009 by the Public Debt Division of the GAO, using its instrument by instrument debt database, were misreported by the GAO in the April and 2 October EDP notifications, where only a negative amount of € 45 mn was included in the EDP tables sent from the GAO to the NSSG, instead of the correct figure of € 495 mn, with the effect of reducing incorrectly the Greek government by € 450 mn. The correct figure was finally introduced in the 21 October EDP notification.
This is to be considered as a case of deliberate misreporting of figures by the GAO in the April and 2 October 2009 EDP notifications.
4.6Debt assumptions and guarantees
Debt assumptions by government linked to the call on guarantees was one of the issues most discussed before 2004 and constituted one of the topics which were the object of the Eurostat 2004 report. At that time, it became clear that rules on debt assumptions had not been followed in the years before 2004, with the effect of incorrectly reducing deficit and debt of government.
On 21 October 2009, the Greek authorities provided Eurostat with revised deficit figures, due to previous misreporting of transactions in debt assumptions originating from calls on guarantees between 2005 and 2008, for amounts of around € 200 mn both in 2007 and 2008 and for much smaller amounts in 2005 and 2006.
In the November 2009 methodological visit, the problem seemed due to additional flows associated to guarantees previously unknown which had been identified as a result of thorough examination of the Treasury Accounts (see paragraph 4.2 above). Nevertheless, although the identification of such flows constitutes an improvement, there are still reasons for concern:
- Noticeable amounts of outstanding guaranteed debt under repeated calls have been identified for the whole EDP reporting period (2005-2008). This might imply that part of this debt, if not all, should be treated as government debt following ESA 95 rules.
- The GAO informed Eurostat on 30 November 2009 that in the event of a call of a guarantee, if it was assumed that a guarantee would be repaid by the beneficiary, nothing was ever recorded in EDP tables, while according to ESA 95 rules in such cases a government expenditure should have been imputed, while if later on a repayment (partial or total) had occurred, government revenue could have been imputed.
In summary, it seems that from 1996 onwards, in spite of repeated assurances provided by the Greek authorities of the contrary, rules were not fully applied in this respect. A full analysis on the amounts of debt assumed by government in the context of guarantee calls might imply a revision of government deficit and debt for past years.
This is to be considered as a case of unwillingness by the GAO to comply with methodological rules in the case of recording of guarantees in the April and 2 October 2009 EDP notifications as well as in the past.
4.7Capital injections
Capital injections in public corporations were one of the issues discussed in the 2004 Eurostat report, as it appeared that in the years before 2004 the rules on capital injections were not applied, with the result that many capital injections which according to rules should have been considered as capital transfers (increasing the deficit of the Greek government) had been considered as financial transactions with no impact on the deficit.
At present, the issue under discussion still concerns, as in 2004 and in previous years, whether capital injections in public enterprises should be considered as financial or non-financial transactions. During the November 2009 methodological visit, it was discovered that some capital injections undertaken in 2008 and in previous years in the public corporations Themis and Tram, up to then considered by the Greek authorities as acquisition of shares without any impact on the deficit, should have been considered as capital transfers, due to the lack of profitability of the two units in question. Moreover, it was also stated by the NSSG that some capital injections undertaken in the past in other corporations, previously considered as acquisition of shares, will also be reclassified as capital transfers, although (unlike in 2004 and in previous years) this time the amount of the revision was not expected to be substantial.
Currently Eurostat does not have any specific information on which will be the corporations involved in the reclassification of capital injections and on the exact amounts involved.
It is likely that the wrong treatment of capital injections, allegedly for small amounts, in the 2009 notifications, but possibly also in past notifications, is due to a lack of information on capital injections and/or on the profitability of enterprises, available to the Greek authorities.
4.8Social security funds
The issue of the continuous considerable surplus of the social security sector figured prominently in the 2004 Eurostat report, where it was described how the data presented by the Greek authorities in the past were unreliable as they were obtained on the basis of an old survey through a number of fragile estimation hypotheses. As a consequence, the surplus of the social security sector had been overestimated by the Greek authorities by 2.8 bn € between 2001 and 2003.
After 2004, the issue of the estimation and compilation of social security data was continuously discussed between Eurostat and the Greek authorities. New more reliable annual and quarterly surveys were put in place.
On 2 October 2009, the surplus of social security was revised downwards by €600 mn19 The considerable decrease of the surplus between the April and 2 October notifications (in previous years such revisions had been smaller) in 2009 led to questions from Eurostat, given the fact that the coverage of social security funds (in terms of revenue and expenditure) in April had been, according to the Greek authorities, already very good at 90% of the total.
The revision, according to the Greek authorities, was due to new questionnaires received between April and October (although the coverage increased only marginally, from 90 to 92% of the total), to incomplete exploitation of the quarterly questionnaires and to the process of unification of some social security funds which seems to have complicated the reporting of data by the surveyed public units.
Eurostat does not find these explanations sufficient and will carry out in the coming months a thorough investigation of the process of calculation by the Greek authorities of the surplus/deficit of the social security sector.
4.9Local government sub-sector
The estimation of deficit/surplus of the local government sub-sector was not among the issues included in the 2004 Eurostat report, although it had been discussed on some occasions before 2004. Until 2005, information on municipalities was gathered via an exhaustive annual questionnaire called the “census”, the results of which were available for EDP notifications only after 3 or 4 years. During recent years, nevertheless, the Greek authorities had also complemented the census via a quarterly survey, which proved useful to increase the reliability of data for local government.
During the November 2009 methodological visit, however, Eurostat was informed that the Greek authorities had decided to stop, or were considering suspending, the quarterly survey, to be replaced by an exhaustive annual questionnaire. As the establishment of a quarterly survey had been decided and agreed in the context of methodological missions, Eurostat strongly advised against such a move before all the statistical and accounting implications of the decision had been properly assessed and a prior discussion with Eurostat had taken place.
There is no open methodological issue at present concerning the recording of deficit/surplus of local government.
4.10Expenditure on military equipment
Expenditure for military equipment by government had been widely underestimated in the years between 1997 and 2003. This was due to the fact that the Greek authorities had stated that they had full information on the value of military equipment delivered, while in reality this was not the case, as information on deliveries was confidential and as a result only a small part of the military equipment delivered was recorded as government expenditure. Following this, it was decided to use cash data for recording expenditures on military equipment, on a temporary basis, to make sure that no expenditure would be left unrecorded.
Starting from the 2007 EDP notification, the Greek authorities are able to comply with ESA 95 rules, recording military expenditure as intermediate consumption at the moment in which the equipment is delivered, as full information on deliveries is now, according to the Greek authorities, available. Data on military expenditures were not revised between the April and the October notifications.
There is no open methodological issue at present concerning the recording of deficit/surplus of military expenditures.
4.11Recording of tax revenues
The recording of taxes was discussed several times with the Greek authorities in the years before 2004 and the issue was included in the 2004 Eurostat report. In particular, up to 2003, the Greek authorities used a system based on assessments and declarations, where the amounts recorded should have been adjusted by a coefficient reflecting amounts unlikely to be collected. Nevertheless, in 2003, Eurostat found out that the coefficients used by the Greek authorities were unrealistically low, and as a consequence the amounts of taxes recorded as revenue for the Greek government had been constantly overestimated. As a result, Eurostat asked the Greek authorities to switch to a method based on time adjusted cash20.
In the 2 October 2009 notification, data on tax revenue were revised downwards by € 650 mn. In the discussion on the issue during the November 2009 methodological visit, it became clear that, contrary to what had been stated by the Greek authorities during the missions undertaken by Eurostat between 2005 and 2008, at the end of March (the time of the first EDP notification of the year), data on direct and indirect tax revenues for the months of January and February of year n (to be imputed as revenue for the year n-1) are simply not available and not known and have therefore to be fully (and not partly) estimated, on the basis of data contained in the Budget Introductory reports of year n and of the annual accrual figures for taxes included in the Stability and Growth Programme prepared by the Macroeconomic Policy Division of the Ministry of Finance. Even though the NSSG had ultimately used more conservative rates of growth for tax revenue compared to the two above mentioned sources, the estimate still proved to be rather optimistic for the months of January and February 2009, leading therefore to a downward revision of tax revenue in 2008.
It is to be underlined, nevertheless, that according to the Greek authorities the issue will be solved in the near future, as all future April EDP notifications, from April 2010 onwards, will be based on full information of tax revenues of the months of January and February.
It is likely that the overestimation of tax revenues in the April 2009 notification is due to methodological errors by the NSSG in the estimation of tax revenues of the months January and February 2009.
4.12Extra-budgetary funds and DEKA
The issue of the deficit/surplus of the 385 existing extra-budgetary funds in Greece, was not one of the objects of the 2004 Eurostat report. However, discussions were often held between Eurostat and the Greek authorities concerning the coverage and reliability of the survey which is the source of information on extra-budgetary funds as well as on DEKA, a unit created in 1997 with the purpose of dealing with enterprises to be privatised, which was reclassified (under instruction from Eurostat) in 2003 as a government unit and the accounts of which have since been included among the extra-budgetary funds. The report of 2004 discussed the issues of the classification of DEKA, of the transactions between DEKA and government units (some of which had been incorrectly treated at the time as revenue of government) and of the capital injections of government in DEKA (incorrectly treated as increases of shares).
In the EDP notification of 21 October 2009, one of the revised items for 2008 compared to the April notification related to an amount received from government by DEKA, as a cash facility to pay its obligations. According to the information received by Eurostat during the November 2009 methodological visit, the revision took place because the new government decided to cancel the loan, treating it instead as government expenditure. Eurostat pointed out that as the government decided to assume the loan in 2009, and not in 2008, the revision of data increasing government deficit should be recorded in 2009 and not in 2008, increasing government deficit in 2009.
However, Eurostat also pointed out that as DEKA is classified since 2003 inside the general government, the transaction had also to be consolidated at the level of general government, and hence the debt cancellation of government should not have had any impact at all. Nevertheless, it also appeared that the amount received by DEKA in 2008 by government might have been incorrectly treated as a financial transaction for government but as revenue for DEKA. Therefore, the correction of 230 mn € increasing government deficit in 2008 may have been correct after all, although for reasons which are different from those on which the Greek authorities based the revision.
It is likely that the wrong recording of flows between government and DEKA in the past is due to methodological errors by the GAO. The issue will have nevertheless to be thoroughly investigated by Eurostat, also in order to make sure that transactions between government and DEKA had been correctly recorded in the years before 2008.
4.13EU grants
The issue of EU grants was prominent in the 2004 Eurostat report. Eurostat had discovered at the time that payments received from the EU in favour of institutional units outside government, when transiting government accounts, were entering them as revenue for government (with a positive impact on government deficit) and exiting them not as expenditure (as would have been appropriate) but as financial transactions, without any impact on the deficit. Following Eurostat's intervention, a correction was undertaken by the Greek authorities, increasing the government deficit. Other irregularities and mistakes were found in the period between 2005 and 2008, and a reservation was expressed in the April 2008 notification due to “the recording of EU grants in 2006 and 2007”.
In the 2 October 2009 notification Eurostat observed a downward revision to the accrual adjustment of EU grants in table 2A for 2008 compared to the April notification (from € 1636 mn to € 1450 mn), decreasing government deficit, and investigated the reasons for this revision. The Greek authorities (the NSSG) answered that the revision was due to new information on the applications submitted for EU grants. In fact, in the 21 October notification the figure was revised up again to € 1666 mn (increasing the deficit of 2008) and it appeared that the correction made by the Greek authorities on 2 October should have never been introduced as the data on applications submitted never changed between the April and October notifications, as stated by the Single Payment Authority.
This is to be considered as a case of deliberate misreporting of figures by the NSSG in the 2 October 2009 EDP notification.
4.14Hospital liabilities
There was no particular issue with respect to public hospitals before 2004 and in the 2004 Eurostat report. The issue of hospitals liabilities started to be discussed in 2005, when it became apparent that considerable amounts related to past expenditures of hospitals (€ 1.3 bn), for the years 2002 to 2004, had in fact never been recorded. It was then decided to record these liabilities in the year in which they had been incurred. At the time the Greek authorities assured Eurostat that the delayed recording of expenditures would not be repeated in the future.
Nevertheless, upon examination of the October notifications, two problems became apparent:
- the NSSG, in spite of a total amount of around € 3.3 bn of unpaid expenditures of hospitals (which should have impacted government deficit) for the period from 2005 to 2008 being reported in the survey on hospitals, recorded only around € 2.3 bn of liabilities in the April and 2 October EDP notifications, incorrectly decreasing the government deficit (compared to the result of the survey) for a minor amount in 2006 and for bigger amounts in the years 2007 and 2008. This is to be considered as a case of deliberate misreporting of figures in the 2007, 2008 and 2 October 2009 EDP notification.
- In the 21 October 2009 notification, an amount of € 2.5 bn was added to the government deficit of 2008 on top of the € 2.3 bn. This was done according to the Greek authorities under a direct instruction from the Ministry of Finance, in spite of the fact that the real total amount of hospital liabilities is still unknown, that there was no justification to impute this amount only in 2008 and not in previous years as well, and that the NSSG had voiced its dissent on the issue to the GAO and to the MOF. This is to be considered as a wrong methodological decision taken by the GAO.
In summary, despite the assurances of the Greek authorities in 2005 that the issue of unrecorded hospitals liabilities in past years (with the consequent ex-post recording of considerable amounts of expenditures in the past) would not have incurred again, the same situation, and possibly for even higher amounts than in 2005, has occurred again in 2009. Beyond the misreporting by the statistical reporting authorities, it emerges that the liabilities of hospital expenditure are not properly recorded by the hospitals themselves, casting severe doubts on the capacity of the Greek authorities to respect accounting rules and on the accountability of Greek institutional framework.
5Conclusions
This report responds to the ECOFIN Council conclusion of 10 November 2009 inviting the Commission to prepare a report on the renewed problems in the Greek fiscal statistics.
The reliability of Greek government deficit and debt statistics has been the subject of continuous and unique attention for several years. In 2004, Eurostat produced a comprehensive report on the revision of the Greek government deficit and debt figures, showing how the Greek statistical authorities had misreported figures on deficit and debt in the years between 1997 and 2003. On five occasions since 2004 reservations have been expressed by Eurostat on the Greek data in the biannual press release on deficit and debt data. When the Greek EDP data have been published without reservations, this has been the result of Eurostat interventions before or during the notification period in order to correct mistakes or inappropriate recording, with the result of increasing the notified deficit. Other elements of this continuous attention are a high number of visits, including four methodological visits, and an action plan agreed with the Greek authorities, addressing the statistical problems that could be diagnosed by Eurostat. That action plan was regularly reviewed by Eurostat. Though eventually an overall level of completion was achieved, given that Eurostat is restricted to statistical matters in its work the measures foreseen in the action plan were mainly of a methodological nature, and did not address the issues of institutional settings, accountability, responsibility and political interference.
The 2004 events led to amendments of the EU legal framework for fiscal data in order to strengthen that framework and to improve the monitoring by the Commission of data provided by Member States in the context of the EDP notifications exercises. The existing legal framework and the governance system for government deficit and debt data at EU level are in general functioning well and produce fiscal data of a generally high quality. It is important to acknowledge the overall efficient and loyal cooperation between national authorities and the Commission that characterises this governance system. The events which have occurred in Greece, as described in this report, are therefore not considered as systemic and relate to individual, country-specific problems.
The most recent revisions are an illustration of the lack of quality of the Greek fiscal statistics (and of Greek macroeconomic statistics in general) and show that the progress in the compilation of fiscal statistics in the country, and the intense scrutiny by Eurostat since 2004, have not sufficed to bring the quality of Greek fiscal data to the level reached by other EU Member States. Even if the existing governance framework for fiscal statistics at EU level functions satisfactorily and enables improvements of a statistical and methodological nature, it cannot prevent deliberate misreporting of data.
The situation described in this report, and in particular what emerged in general from the information gathered by Eurostat during the November 2009 methodological visit and from the ensuing correspondence with the Greek authorities, can be summarised as follows:
1) Regarding government deficit and debt data for Greece: for the time being, in view of the described size and nature of the data revisions, the lack of reliability and the shortage of evidence supporting the deficit figures reported, Eurostat is at present not in a position to validate figures which are of acceptable statistical quality.
2) Regarding institutional issues, actors and procedures, the present institutional setting, which has showed its weakness, inefficiency and permeability to political interference throughout the years, does not guarantee the professional independence and full accountability of the NSSG and of the other services involved in the domain of EDP data.
3) The following findings should be mentioned:
- Severe irregularities in the EDP notifications of April and October 2009, including unreliability of data, non-respect of accounting rules, and timing of the notification;
- Poor cooperation between the national services involved in the compilation of EDP figures and lack of independence of the NSSG and the GAO;
- An institutional setting and a public accounting system inappropriate for a correct reporting of EDP statistics, notably non-transparent or improperly documented bookkeeping;
- Lack of accountability in the individual provision of figures used in EDP notification, such as, in some cases, absence of written documentation or certification.
- Unclear responsibility and/or lack of responsibility of the national services providing source data or compiling statistical data, with inappropriate adjustments to data, combined with unclear empowerment of officials responsible for the data.
More specifically, with regard to compilation issues, the investigation of Eurostat has revealed significant weaknesses in the national services principally responsible for EDP data: the National Statistical Service of Greece (NSSG), the General Accounting Office (GAO) and the Ministry of Finance (MOF).
From a more general perspective, although the responsibility for monitoring the fiscal data is vested in the Commission, it does not directly compile government data for the Member States. The Commission depends greatly on the data compiled and reported by Member States, as well as on the administrative ability, good will and co-operation of the respective national authorities; moreover, it does not posses audit powers.
In this context, the shortcomings in the operational and administrative capacity of some of the institutions involved in the production of EDP statistics in Greece, referred to in this report, will have to be decisively addressed by the Greek authorities without further delay. Unless the institutional weaknesses uncovered during the investigation of the irregularities underlying the 2009 notifications of data are corrected and proper checks and balances introduced, the reliability of Greek deficit and debt data will remain in question.
The Commission is fully committed to continue cooperating with the Greek authorities with a view to supporting their efforts to improve the collection and processing of government statistics in order to address the recurrent shortcomings and restore the confidence in Greek statistics.
List of Annexes
Annex 1 – Methodological issues
Annex 2 – Quarterly non-financial and financial government statistics
Annex 3 – Reservations on the quality of the Greek data expressed by Eurostat between 2005 and 2009
Annex 4 – Action plan/list of recommendations as implemented by December 2009
Annex 5 – Exchange of letters on hospital liabilities
Annex 6 – Infringement procedure
1 According to the appropriate regulations and practices, the planned deficit and debt data reported by Member States are not scrutinised by Eurostat or any other Commission service, and the Commission publishes its own macroeconomic forecasts. Hence, although there is an intrinsic link between the quality of statistics (i.e. data for a past year) and the reliability of planned data (i.e. estimates or forecasts for an ongoing or future period), this report does not elaborate on the technical procedures or institutional issues related to the preparation and dissemination of planned data.
2 The Council REGRETS the renewed problems in the Greek fiscal statistics. The Council CALLS ON the Greek government to urgently take measures to restore the confidence of the European Union in Greek statistical information and the related institutional setting. The Council INVITES the Commission to produce a report before the end of 2009. Moreover, the Council INVITES the Commission to propose the appropriate measures to be taken in this situation. In this context, the Council WELCOMES the commitment by the Government to address this issue swiftly and seriously and CONSIDERS the measures announced recently, such as those aiming to make the National Statistical Service fully independent, to be steps in the right direction.
3 Prior to that Regulation (EC) 322/97 stressed the principles of impartiality, reliability, relevance, cost-effectiveness, statistical confidentiality and transparency
4 Communication from the Commission to the European Parliament and to the Council on the independence, integrity and accountability of the national and Community statistical authorities (COM 2005 217 of 25.5.2005)).
5 Council Regulation (EC) No 2223/96 of 25 June 1996 on the European system of national and regional accounts in the Community (OJ L 310, 30.11.1996, p.
1).
6 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 (Codified version) (OJ L 145, 10.6.2009, p.
1).
7 The proposed amendments in COM(2005)71 sought to increase the transparency of the EDP-related statistics, and to strengthen the powers of Eurostat with respect to data quality. In particular, the Commission sought to establish 'in-depth monitoring visits'. Coupled with a general requirement for Member States to promptly provide the Commission (Eurostat) with access to the information required for the purposes of the data quality assessment, this would have given Eurostat audit-like powers.
See www.cc.cec/sg_vista/cgi-bin/repository/getdoc">www.cc.cec/sg_vista/cgi-bin/repository/getdoc
8 See eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2005:337:0001:0006:EN:PDF
9 The Bank also has a reporting responsibility to the ECB under the ECB Guideline of 31 July 2009 on government finance statistics (GFS) with the data to be used for economic and monetary analysis. The GFS required by the ECB cover the following information for general government: (i) revenue and expenditure including deficit/surplus by subsectors (central, local, social security), selected data on EU flows and some aggregates, (ii) deficit-debt adjustments statistics and (iii) debt statistics. The data must comply with the principles and definitions of ESA95. Explanations of the reasons of the revisions of a significant magnitude should be provided to the ECB by the responsible institution (which is the BOG in Greece). See ECB/2009/200; 2009/627/EC.
10 In October 2007, a peer review on the statistics in general was carried out in Greece as part of the first wave of peer reviews in all NSIs / Eurostat. The review recommended the National Statistical Service of Greece (NSSG) to substantially strengthen the institutional set up; improvement action was agreed with the NSSG top management.
11 epp.eurostat.ec.europa.eu/portal/pls/portal/!PORTAL.wwpob_page.show?_docname=1758246.PDF
12 Greece: Due to ongoing discussions with the statistical authorities in Greece, notably on the surplus of social security funds, the notified figures for debt and deficit are to be considered as provisional and could be revised.
13 The Council Regulation 3605/93 provided for EDP notifications before 1 March and 1 September.
14 Greece:
The revision of data for the deficit between the March and September 2004 notifications was carried out on the basis of new information provided by the Greek Authorities, at the request of Eurostat, for the period 2000-2003.
The change in the deficit figure is due mainly to:
- Downward revision for 2003 of estimate of tax revenues (mainly VAT) in public accounts;
- Downward revision for 2003 of payments received from EU institutions in the context of certain structural fund
programmes;
- Reclassification for 2003 of a payment from the postal savings bank to government as a financial transaction;
(for these first three items see News Release 62/2004 of 7 May 2004)
- Under-recording of military expenditures between 2000 and 2003;
- Over-estimation of surplus of social security funds between 2001 and 2003;
- Under-recording of interest between 2000 and 2003.
The revision of data for the debt between the March and September 2004 notifications was carried out on the basis of new information provided by the Greek Authorities for the period 2000-2003. The change in the debt figure is due mainly to:
- Under-estimation of outstanding debt, notably in relation to bonds with capitalised interest;
- Over-estimation of consolidating assets of social security.
15 On Eurostat Website: Reports of methodological visit to Greece in 2008
16 On Eurostat Website: Reports of methodological visit to Greece in 2008
17 Eurostat was informed that the balances of some of these accounts appeared both in the accounts of the Ministries and the balance sheets of the public entities 'through which the different projects are carried out'. Eurostat asked the Greek authorities whether this might imply a double counting of these accounts in the previous notifications.
18 In his correspondence with Eurostat, the General Secretary of the NSSG claimed to have initially received from the GAO new figures for the EDP notification, on the basis of which the deficit for 2008 was 6.7% of GDP; to have been subsequently instructed not to send the new EDP notification to Eurostat until after a meeting of the GAO with the Minister; and finally to have received instructions, on 21 October, to send a new EDP notification to Eurostat to replace the one sent on 2 October, whatever the objections the NSSG could have about the new figures.
19 the revision would in fact have been bigger if Eurostat had not advised the Greek authorities to allocate half of the positive discrepancy of the social security to the reduction of the surplus of the sector
20 Tax revenues for year n are calculated taking into account tax revenues obtained from March of year n until February of year n+1.
30