Not the law itself: Notice

Address https://gelex.eu/eli/EU/notice/oj/C/2016/375/01@2016-10-13/en

Contents

1. Preamble
2. GENERAL INTRODUCTION
3. CHAPTER 1 The statement of assurance and supporting information
4. (b)
5. ANNEX
6. ANNEX 1.1 AUDIT APPROACH AND METHODOLOGY
7. PART 1 Audit approach and methodology for the reliability of accounts 2.
8. A.
9. B.
10. PART 2 Audit approach and methodology for the regularity of transactions 3.
11. How we test transactions 4.5.6.7.8.9.10.
12. How we evaluate and present the results of transaction testing 11.12.13.14.
13. Estimated level of error 15.16.17.
14. How we examine systems and report the results 18.19.
15. How we arrive at our opinions in the statement of assurance 20.
16. Suspected fraud 21.
17. PART 3 Link between the audit opinions for the reliability of accounts and the regularity of transactions 22.23.
18. ANNEX 1.2 1.2.3.4.5.6. Differences and similarities between the US and the ECA models for estimating the level of error in budgetary spending US AND EU APPROACHES TO TESTING AND REPORTING ON IMPROPER AND IRREGULAR PAYMENTS
19. CHAPTER 2 Budgetary and financial management
20. CHAPTER 3 Getting results from the EU budget
21. ANNEX 3.2 FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR PERFORMANCE ISSUES
22. CHAPTER 4 Revenue
23. (a)
24. ANNEX 4.1 RESULTS OF TRANSACTION TESTING FOR REVENUE
25. ANNEX 4.2 FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR REVENUE
26. ANNEX 4.3 NUMBER OF OUTSTANDING GNI RESERVATIONS, VAT RESERVATIONS AND TOR OPEN POINTS BY MEMBER STATE AT 31.12.2015
27. CHAPTER 5 Competitiveness for growth and jobs
28. ANNEX 5.1 RESULTS OF TRANSACTION TESTING FOR COMPETITIVENESS FOR GROWTH AND JOBS
29. ANNEX 5.2 FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR COMPETITIVENESS FOR GROWTH AND JOBS
30. CHAPTER 6 Economic, social and territorial cohesion
31. pt_11
32. pt_12
33. ANNEX 6.1 RESULTS OF TRANSACTION TESTING FOR ECONOMIC, SOCIAL AND TERRITORIAL COHESION
34. ANNEX 6.2 FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR ECONOMIC, SOCIAL AND TERRITORIAL COHESION
35. CHAPTER 7 Natural resources
36. ANNEX 7.1 RESULTS OF TRANSACTION TESTING FOR NATURAL RESOURCES
37. ANNEX 7.2 3.46.4.37.4.38.3.37.4.44. FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR NATURAL RESOURCES
38. CHAPTER 8 Global Europe and Security and citizenship
39. ANNEX 8.1 RESULTS OF TRANSACTION TESTING FOR GLOBAL EUROPE
40. ANNEX 8.2 FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR GLOBAL EUROPE
41. CHAPTER 9 Administration
42. ANNEX 9.1 RESULTS OF TRANSACTION TESTING FOR ADMINISTRATION
43. ANNEX 9.2 FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR ADMINISTRATION

In accordance with the provisions of Article 287(1) and (4) of the TFEU and Articles 148(1) and 162(1) of Regulation (EU, Euratom) No 966/2012 of the European Parliament and of the Council of 25 October 2012 on the financial rules applicable to the general budget of the Union and repealing Council Regulation (EC, Euratom) No 1605/2002 and Articles 43, 48 and 60 of Council Regulation (EC) No 215/2008 of 18 February 2008 on the Financial Regulation applicable to the 10th European Development Fund, as amended by Regulation (EU) No 567/2014 the Court of Auditors of the European Union, at its meeting of 14 July 2016, adopted its ANNUAL REPORTS concerning the financial year 2015. The reports, together with the institutions' replies to the Court's observations, were transmitted to the authorities responsible for giving discharge and to the other institutions. The Members of the Court of Auditors are: Vítor Manuel da SILVA CALDEIRA (President), Kersti KALJULAID, Karel PINXTEN, Henri GRETHEN, Szabolcs FAZAKAS, Louis GALEA, Ladislav BALKO, Lazaros S. LAZAROU, Hans Gustaf WESSBERG, Pietro RUSSO, Ville ITÄLÄ, Kevin CARDIFF, Baudilio TOMÉ MUGURUZA, Iliana IVANOVA, George PUFAN, Neven MATES, Alex BRENNINKMEIJER, Danièle LAMARQUE, Nikolaos MILIONIS, Phil WYNN OWEN, Klaus-Heiner LEHNE, Oskar HERICS, Bettina JAKOBSEN, Janusz WOJCIECHOWSKI, Samo JEREB, Jan GREGOR, Mihails KOZLOVS, Rimantas ŠADŽIUS.

GENERAL INTRODUCTION

Txt. 1. The European Court of Auditors is the institution established by the Treaty on the Functioning of the European Union to carry out the audit of European Union (EU) finances. As the EU’s external auditor we act as the independent guardian of the financial interests of the citizens of the Union and contribute to improving EU financial management. More information on our work can be found in our annual activity report, our special reports on EU spending programmes and revenue, our landscape reviews and our opinions on new or updated EU laws or on other decisions with financial management implications. This is our 39th annual report on the implementation of the EU budget and covers the 2015 financial year. A separate annual report covers the European Development Funds. The general budget of the EU is approved annually by the Council and the European Parliament. Our annual report, together with our special reports, provides a basis for the discharge procedure, in which the European Parliament, after a recommendation from the Council, decides whether the Commission has satisfactorily carried out its responsibilities for implementing the budget. On publication we forward our annual report to national parliaments, the European Parliament and the Council. The central part of our annual report is the statement of assurance on the reliability of the consolidated accounts of the EU and on the legality and regularity of transactions (referred to in the report as regularity of transactions). We supplement this statement with specific assessments of each major area of EU activity. The chapter structure within the report reflects the headings of the multiannual financial framework (MFF). This framework came into force in 2014, and is intended to determine the size and distribution of EU spending over the period until 2020. There are no separate financial statements for individual MFF headings and therefore conclusions presented in specific assessments do not constitute audit opinions. Instead, these chapters present significant issues specific to each MFF heading. Our report this year is organised as follows: (1) chapter 1 contains the statement of assurance and a summary of the results of our audit on the reliability of accounts and on the regularity of transactions; (2) chapter 2 presents our findings on budgetary and financial management; (3) chapter 3 contributes to our strategic objective of improving EU public accountability and focusses on the extent to which the performance framework in the area of Horizon 2020, its research and innovation programme, allows the Commission to monitor and steer progress towards its objectives; (4) chapter 4 presents our findings on revenue; (5) chapters 5 to 9 show, for the main MFF headings, the results of our testing of the regularity of transactions and examination of annual activity reports and other elements of internal control systems, and other governance arrangements. Txt. 2. The Commission’s replies (or replies of other EU institutions and bodies, where appropriate) to our observations are presented within the report. Our findings and conclusions take into account these replies. However, it remains our responsibility, as external auditor, to report our audit findings, to draw conclusions from those findings, and thus to provide an independent and impartial assessment of the reliability of the accounts and the regularity of transactions.

The European Court of Auditors is the institution established by the Treaty on the Functioning of the European Union to carry out the audit of European Union (EU) finances. As the EU’s external auditor we act as the independent guardian of the financial interests of the citizens of the Union and contribute to improving EU financial management. More information on our work can be found in our annual activity report, our special reports on EU spending programmes and revenue, our landscape reviews and our opinions on new or updated EU laws or on other decisions with financial management implications. This is our 39th annual report on the implementation of the EU budget and covers the 2015 financial year. A separate annual report covers the European Development Funds. The general budget of the EU is approved annually by the Council and the European Parliament. Our annual report, together with our special reports, provides a basis for the discharge procedure, in which the European Parliament, after a recommendation from the Council, decides whether the Commission has satisfactorily carried out its responsibilities for implementing the budget. On publication we forward our annual report to national parliaments, the European Parliament and the Council. The central part of our annual report is the statement of assurance on the reliability of the consolidated accounts of the EU and on the legality and regularity of transactions (referred to in the report as regularity of transactions). We supplement this statement with specific assessments of each major area of EU activity. The chapter structure within the report reflects the headings of the multiannual financial framework (MFF). This framework came into force in 2014, and is intended to determine the size and distribution of EU spending over the period until 2020. There are no separate financial statements for individual MFF headings and therefore conclusions presented in specific assessments do not constitute audit opinions. Instead, these chapters present significant issues specific to each MFF heading. Our report this year is organised as follows:

chapter 1 contains the statement of assurance and a summary of the results of our audit on the reliability of accounts and on the regularity of transactions;

chapter 2 presents our findings on budgetary and financial management;

chapter 3 contributes to our strategic objective of improving EU public accountability and focusses on the extent to which the performance framework in the area of Horizon 2020, its research and innovation programme, allows the Commission to monitor and steer progress towards its objectives;

chapter 4 presents our findings on revenue;

chapters 5 to 9 show, for the main MFF headings, the results of our testing of the regularity of transactions and examination of annual activity reports and other elements of internal control systems, and other governance arrangements.

The Commission’s replies (or replies of other EU institutions and bodies, where appropriate) to our observations are presented within the report. Our findings and conclusions take into account these replies. However, it remains our responsibility, as external auditor, to report our audit findings, to draw conclusions from those findings, and thus to provide an independent and impartial assessment of the reliability of the accounts and the regularity of transactions.

CHAPTER 1 The statement of assurance and supporting information

Txt. 1. THE COURT'S STATEMENT OF ASSURANCE PROVIDED TO THE EUROPEAN PARLIAMENT AND THE COUNCIL — INDEPENDENT AUDITOR’S REPORT We have audited: (a) the consolidated accounts of the European Union which comprise the consolidated financial statements and the aggregated reports on the implementation of the budget for the financial year ended 31 December 2015, approved by the Commission on 11 July 2016, and (b) the legality and regularity of the transactions underlying those accounts, as required by Article 287 of the Treaty on the Functioning of the European Union (TFEU). Txt. 2. Management's responsibility In accordance with Articles 317 to 318 of the TFEU and the Financial Regulation, the Commission is responsible for the preparation and fair presentation of the consolidated accounts of the European Union on the basis of internationally accepted accounting standards for the public sector and for the legality and regularity of the transactions underlying them. This responsibility includes the design, implementation and maintenance of internal control relevant to the preparation and presentation of financial statements that are free from material misstatement, whether due to fraud or error. The Commission bears the ultimate responsibility for the legality and regularity of the transactions underlying the accounts of the European Union (Article 317 of the TFEU). Auditor's responsibility Our responsibility is to provide, on the basis of our audit, the European Parliament and the Council with a statement of assurance as to the reliability of the accounts and the legality and regularity of the transactions underlying them. We conducted our audit in accordance with the IFAC International Standards on Auditing and Codes of Ethics and the INTOSAI International Standards of Supreme Audit Institutions. These standards require that we comply with ethical requirements and plan and perform the audit to obtain reasonable assurance as to whether the consolidated accounts of the European Union are free from material misstatement and the transactions underlying them are legal and regular. An audit involves performing procedures to obtain audit evidence about the amounts and disclosures in the consolidated accounts and the legality and the regularity of the transactions underlying them. The procedures selected depend on the auditor's judgment, including an assessment of the risks of material misstatement of the consolidated accounts and of material non-compliance of the underlying transactions with the requirements of the legal framework of the European Union, whether due to fraud or error. In making those risk assessments, internal control relevant to the preparation and fair presentation of the consolidated accounts and legality and regularity of underlying transactions, is considered in order to design audit procedures that are appropriate in the circumstances, but not for the purpose of expressing an opinion on the effectiveness of internal control. An audit also includes evaluating the appropriateness of accounting policies used and the reasonableness of accounting estimates made, as well as evaluating the overall presentation of the consolidated accounts. For revenue, our examination of value added tax and gross national income-based own resources takes as its starting point the relevant macroeconomic aggregates on which these are calculated, and assesses the Commission's systems for processing these until the contributions of the Member States have been received and recorded in the consolidated accounts. For traditional own resources, we examine the accounts of the customs authorities and analyse the flow of duties until the amounts are received by the Commission and recorded in the accounts. For expenditure, we examine payment transactions when expenditure has been incurred, recorded and accepted. This examination covers all categories of payments (including those made for the purchase of assets) other than advances at the point they are made. Advance payments are examined when the recipient of funds is required to provide or provides justification for their proper use and the advance payment is cleared or becomes recoverable. We consider that the audit evidence obtained is sufficient and appropriate to provide a basis for our opinions. Reliability of the accounts Opinion on the reliability of the accounts In our opinion, the consolidated accounts of the European Union for the year ended 31 December 2015 present fairly, in all material respects, the financial position of the Union as at 31 December 2015, the results of its operations, its cash flows, and the changes in net assets for the year then ended, in accordance with the Financial Regulation and with accounting rules based on internationally accepted accounting standards for the public sector. Legality and regularity of the transactions underlying the accounts Revenue Opinion on the legality and regularity of revenue underlying the accounts In our opinion, revenue underlying the accounts for the year ended 31 December 2015 is legal and regular in all material respects. Payments Basis for adverse opinion on the legality and regularity of payments underlying the accounts Expenditure recorded in 2015 under the multi-annual financial framework headings covering operational spending is materially affected by error. Our estimated level of error for payments underlying the accounts is 3,8 %. Our overall conclusion is driven by the higher estimated level of error for spending on a reimbursement basis and is corroborated by the Commission’s analysis of amounts at risk presented in the annual management and performance report for the EU budget. Adverse opinion on the legality and regularity of payments underlying the accounts In our opinion, because of the significance of the matters described in the basis for adverse opinion on the legality and regularity of payments underlying the accounts paragraph, the payments underlying the accounts for the year ended 31 December 2015 are materially affected by error. Other information The Financial Statement Discussion and Analysis is not a part of the financial statements. The information given in the Financial Statement Discussion and Analysis is consistent with the financial statements. 14 July 2016 Vítor Manuel da SILVA CALDEIRA President European Court of Auditors 12, rue Alcide De Gasperi, 1615 Luxembourg, LUXEMBOURG THE COURT’S OBSERVATIONS INTRODUCTION EU spending is a significant tool for achieving policy objectives EU spending amounted to 145,2 billion euro in 2015. It represents a significant instrument for achieving policy objectives. Other significant instruments are the use of legislation and the freedom for goods, services, capital and people to move throughout the Union. EU spending represents 2,1 % of EU Member States’ (see Figure 1.1) total general government expenditure, and 1,0 % of EU gross national income. As set out in the Treaty, we: (5) check whether the EU budget is used in accordance with applicable laws and regulations; (6) give our opinion on the accounts of the EU; (7) report on whether EU spending is economic, efficient and effective; and (8) advise on proposed legislation with a financial impact. Txt. 3. The work we do for the statement of assurance fulfils the first and second of these objectives. It helps us identify risks affecting the economy, efficiency and effectiveness of spending. It provides a key input to our opinions on proposed financial legislation. 2015 EU spending in each Member State as a share of gross national income (GNI) and general government expenditure Member States GNI: Agreed set of forecasts of traditional own resources and VAT/GNI bases — 19.5.2016 — European Commission. Member States — general government expenditure — Eurostat — annual national accounts. See Annex 1.3 for explanation of country codes. Covers the expenditure, salaries and other costs of EU bodies located in the Member States. Source: European Union spending (European Commission accounting data, compiled by ECA). THE COURT’S OBSERVATIONS This chapter of the annual report: (9) sets out the background to our statement of assurance and summarises our audit approach (see Annex 1.1), findings and conclusions; (10) discusses key audit matters arising from the audit of the Union’s financial statements; (11) compares our results on the regularity of spending in 2015 with findings in recent years and with the estimates of the European Commission. (12) Annex 1.2 discusses United States (US) and EU approaches to testing and reporting on improper and irregular payments. Txt. 4. THE ACCOUNTS WERE NOT AFFECTED BY MATERIAL MISSTATEMENTS Our observations concern the consolidated accounts (the accounts) of the European Union for the financial year 2015. We received them, together with the accounting officer’s letter of representation, on 12 July 2016. The accounts are accompanied by a Financial Statement Discussion and Analysis. This analysis is not covered by our audit opinion. In accordance with the auditing standards, we have, however, assessed its consistency with information in the accounts. The accounts published by the Commission show that, at 31 December 2015, total liabilities amounted to 226,1 billion euro compared to 153,7 billion euro of total assets. The deficit for 2015 amounts to 13,0 billion euro. Our audit found that the accounts were not affected by material misstatements. We present our observations on the financial and budgetary management of EU funds in Chapter 2. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES In 2015 the Commission further strengthened alignment with international standards and with the Recommended Practice Guideline for preparing and presenting Financial Statement Discussion and Analysis. KEY AUDIT MATTERS Key audit matters are those matters that, in our professional judgment, were of most significance in our audit of the financial statements of the current period. These matters were addressed in the context of our audit of the financial statements as a whole, and in forming our opinion thereon, but we do not provide a separate opinion on these matters. We sought additional information from the Commission to support the valuation of financial instruments under shared management Authorities in the Member States transfer a part of the funding advanced by the Commission into financial instruments in the form of loans, equity instruments or guarantees. For the 2007-2013 MFF, EU law did not require these authorities to produce periodic reports on sums held in these instruments in time for preparation of the accounts. The Commission therefore estimated the use made of these advances, assuming that funds would be used in full and evenly over the period of operation (initially up to 31 December 2015 but later extended — without legal authority — to 31 March 2017). By the end of 2014, of the 16 billion euro (including national contributions) transferred into cohesion financial instruments under shared management, only 9,2 billion euro (57 %) had been passed to projects. The Commission recalls its disagreement with the Court’s statement concerning the extension of the period of operation for financial instruments under shared management (see also the Commission’s replies to paragraphs 6.41 and 6.45). THE COURT’S OBSERVATIONS We required the Commission to confirm the validity of its methodology for estimating the amount presented in its balance sheet (3,3 billion euro at 31 December 2015). The Commission therefore launched a data collection exercise, targeted at the Member States presenting the highest balances. We reviewed the procedure put in place for collecting the data from the Member States concerned and taking into consideration the results obtained, we conclude that the overall amount of pre-financing related to financial instruments presented in the balance sheet is fairly stated. We assessed the accounting treatment of the European Fund for Strategic Investments (EFSI) guarantee operations In 2015 the European Fund for Strategic Investments (EFSI) was launched. EFSI is a joint initiative launched by the EU and the EIB to encourage investment across the EU, by increasing the risk bearing capacity of the EIB Group. Under this initiative, the EU budget will provide a guarantee of up to 16 billion euro to cover potential losses on EFSI operations (12,3 billion euro for a debt portfolio and 3,7 billion euro for an equity portfolio) as explained in paragraphs 2.32-2.38. We focused on this area because the set-up of EFSI has given rise to a number of complex accounting issues resulting from the arrangements between the EU and the EIB. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Under the EFSI debt portfolio, the EU budget guarantee covers the first loss of a portfolio of financing operations entered into by the EIB. The Commission therefore accounts for it as a financial guarantee liability. Under the EFSI equity portfolio, the EIB makes equal, parallel investments at its own risk and at the risk of the EU. For the equity investments made by the EIB at the risk of the EU, the EU budget bears the whole risk and each year end the Commission transfers to the EIB a sum equal to the negative value adjustments (unrealised losses), realised losses at disinvestment and EIB funding costs, net of receipts from interests, dividends and realised gains. There is no transfer from the EIB to the EU for unrealised gains above the initial value of the investment. Equity portfolio investments (concerning only the Infrastructure and Innovation Window) covered by the EU budget guarantee amounted to 7,6 million euro at year end. The Accounting Officer exercised significant judgement when assessing and concluding that the EFSI did not meet the criteria for classification as a joint operation. Other complex accounting issues relate to the recognition and measurement of the financial guarantee liability for the debt portfolio, the EU guarantee of equity investments and the recognition of potential gains under the equity portfolio. The Accounting Officer of the Commission exercised, as he is required to do frequently for many other accounting issues, his professional judgement when taking a decision on the accounting treatment for EFSI. This judgement was based on an in-depth analysis of the facts by the accounting services of the Commission and the unanimous support of the EU advisory group on accounting standards. We focused audit work on analysing the control indicators. Had the Commission treated EFSI as a joint operation, this would require some additional disclosures on ownership interests and voting rights, inclusion in note 9 of the financial statements (scope of consolidation), and recognition of an EU share of the assets, liabilities, costs and revenues of the equity portfolio. However, at the end of 2015, this would have had no impact on the net position of the EU. The Commission has included in the notes to the 2015 accounts all mandatory disclosures required by the EU accounting rules in force for these accounts. In addition to this, both the notes and the FSDA provide extensive supplementary information on the nature, extent and financial effects of the EFSI for the EU. For the debt portfolio, we also reviewed the adequacy of the accounting policies and found the accounting treatment to be consistent and in line with international standards. For the equity portfolio we consider that if future years show an increase in the value of equity investments above their initial valuation this should be recognised in the EU accounts. At the end of 2015, the sums involved (see paragraph 1.13) do not have a material impact on the EU consolidated accounts. In the future, the impact of any gains on the net position of the EU would be the same whether the Commission accounts for these operations as a joint operation or a derivative. While we will keep this issue under review, we have concluded that it does not have an impact on our audit opinion for 2015. The Commission welcomes that the Court finds the accounting treatment of the debt portfolio to be consistent and in line with international standards. Discussions on the equity portfolio, which was not implemented in 2015, are still ongoing between the Commission and the EIB Group. The accounting treatment of the equity operations will be established by the Accounting Officer after consulting the EU Advisory Group of Experts on Accounting Standards once the amended legal basis has been finalised. We assessed the increase in accrued charges presented in the accounts In the final accounts of 2015, the Commission estimated that incurred eligible expenses due to the beneficiaries but not yet reported to it amounted to 106 billion euro. It recorded these as accrued expenses. This represents an increase of 18 billion euro on 2014. The Commission will use the experience with the estimation of incurred eligible expenses to further improve the cut-off exercise in the future. We examined the methodology and control systems applied in the most significant directorates-general to the year-end estimates and the cost statements submitted by Member States in the first quarter of 2016. We drew samples on invoices and pre-financings and carried out work on these elements to address the risk that the accrual was misstated. We sought additional explanation from the accounting officer on the increase in the accrued charges as presented in the accounts. The work we performed on the provisional accounts and the net adjustments of 3,3 billion euro made in the final accounts enable us to conclude that the overall amount of accrued charges in the final consolidated balance sheet is fairly stated. THE COURT’S OBSERVATIONS REGULARITY OF TRANSACTIONS: ALTHOUGH OUR ESTIMATED LEVEL OF ERROR REMAINS MATERIAL, THERE WERE SOME AREAS OF IMPROVEMENT We examine EU revenue and spending to assess whether it is in accordance with applicable laws and regulations. We present our audit results on revenue in chapter 4 and on expenditure in chapters 5 to 9 (see Figure 1.2). Pt_a_. 2. Revenue was free from material error (see paragraph 4.22). (i) Spending on MFF heading 5 (Administration) was free from material error (see paragraph 9.16). (ii) Operational spending was affected by material error (see paragraphs 5.35 to 5.36, 6.76 to 6.77, 7.64 to 7.65 and 8.33 to 8.34) with much higher levels of error affecting expenditure driven by the reimbursement of costs (see paragraph 1.22). Pt_b_. 2. In expenditure, we continue to find a material level of error. The estimated level of error in expenditure was 3,8 %, a lower level than in 2014 (4,4 %) (see Figure 1.2). Txt. 5. 2015 Summary of audit results on regularity of transactions Source: European Court of Auditors. Annual report chapter MFF headings Transactions subject to audit (billion euro) Estimated level of error 2015 (%) Confidence interval (%) Estimated level of error 2014 (%) Lower error limit (LEL) Upper error limit (UEL) Competitiveness Heading 1.a 14,5 4,4 2,0 6,7 5,6 Cohesion Heading 1.b 53,9 5,2 2,8 7,6 5,7 Natural Resources Heading 2 58,6 2,9 1,7 4,2 3,6 Global Europe Heading 4 6,9 2,8 1,0 4,6 2,7 Administration Heading 5 9,0 0,6 0,0 1,2 0,5 Other Heading 3, 6 and other 3,0 — — — — Total 145,9 3,8 2,7 4,8 4,4 Revenue 153,8 0 0 0 0 THE COURT’S OBSERVATIONS Our 2015 audit results show a broadly similar picture to recent years Our overall estimated level of error, although improved on recent years, continues to exceed significantly our benchmark for materiality (see Figure 1.3). The estimated level of error (2013 to 2015) Source: European Court of Auditors (see Figure 1.2). THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Reimbursement of costs is linked to higher levels of error EU spending is dominated by two types of expenditure programmes. The differences between two bases for payment (cost reimbursement and entitlement) explain a large part of the variation in the estimated level of error. 1.21. (15) Cost reimbursement; where the EU reimburses eligible costs for eligible activities. They include research projects (chapter 5), training schemes (chapter 6), as well as projects in regional and rural development (chapters 6 and 7) and development projects (chapter 8). In these schemes, recipients must provide information supporting their claim to be engaged in an activity eligible for support and to have incurred costs for which they are entitled to reimbursement. The requirement to provide accurate information gives rise to complicated eligibility rules (what can be claimed) and formal requirements (how to incur costs, e.g. public procurement or state aid rules). Txt. 6. The Commission considers that the impact of the complicated eligibility rules referred to by the Court is significantly mitigated when simplified cost options are used. (16) Entitlement; where payment is based on meeting certain conditions. Such programmes include student and research fellowships (chapter 5), budget support (chapter 8), salaries and pensions (chapter 9). Direct aid for farmers (chapter 7) is based on entitlement and calculated on the basis of information confirmed or declared by farmers; the risk linked to potentially incorrect declarations is largely mitigated by the simpler nature of the information expected from beneficiaries, which can, in large part, be verified before payment through, for example, comparisons with databases and orthophotos. Txt. 7. THE COURT’S OBSERVATIONS Eligibility errors in cost reimbursement schemes dominate the errors detected for 2015. (1) ineligible costs included in cost claims; (2) ineligible projects, activities and beneficiaries; and (3) serious infringements of public procurement rules (Figure 1.4). Pt_a_. 3. For reimbursement expenditure the estimated level of error is 5,2 % (2014: 5,5 %). Typical errors in this area include: Pt_b_. 3. For entitlement expenditure, the estimated level of error is 1,9 % (2014: 2,7 %). Typical errors in this area include small over-declarations by farmers of farm and field areas (paragraph 7.16). Txt. 8. Breakdown of overall estimated level of error by type The 2014 comparator for Errors by Commission and intermediary bodies is the total of Administrative errors in natural resources and Other error types. Source: European Court of Auditors. THE COURT’S OBSERVATIONS Figure 1.5 is based on our transaction testing of EU spending over the last two years for the annual reports on the general budget and the European Development Funds. In general, the more complex the information that is required from beneficiaries, the higher the risk of error. The relationship between basis for payment and estimated level of error in EU transactions (2014-2015) Reimbursement for Global Europe includes multi-donor projects which in practice have many of the attributes of entitlement spending and are affected by lower levels of error. Source: European Court of Auditors. THE COURT’S OBSERVATIONS We analysed other risk factors and their relationship with the estimated level of error. We found that within direct management spending two risk factors have a significant impact on the level of error. (19) In areas where the rules provide for an audit or inspection before the payment is made we find, in general, a lower level of error. (20) Reimbursements to projects involving a number of partners (e.g. consortia) have higher levels of error than reimbursements to a single beneficiary. Txt. 9. Management mode has a limited impact on levels of error The different risk patterns of reimbursement and entitlement schemes, rather than management modes, influences the level of errors in different areas of spending.The estimated level of error shows a much stronger relationship with the basis for payment than with the management mode. We found the highest levels of error over the last two years in Competitiveness (managed directly by the Commission and indirectly through entrusted entities) and Cohesion (under shared management). Reimbursement schemes dominate spending in both of these areas (see Figure 1.6). Average estimated level of error by management mode and by basis for payment (2014-2015) (1) The two sides to the graph present the same amount of expenditure. On the left hand side spending is split on the basis of management mode, showing that the great majority of operational spending is under shared management. The right hand side shows spending split by the basis for payment and shows that reimbursement spending is significantly greater than entitlement spending. The vertical position of the bars indicates the estimated level of error for each. This shows the limited impact of management mode and the significant impact of basis for payment. Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Results in different areas of spending show distinct patterns of error As shown in Figure 1.7, Cohesion was the biggest contributor to our estimated level of error for 2015, followed by Natural Resources, Competitiveness and Global Europe. This distribution is in line with our findings for 2014. The Commission shares the Court’s assessment of the errors and will duly follow them up to protect the EU budget. Contribution of different MFF headings to the overall estimated level of error (2015) Source: European Court of Auditors. THE COURT’S OBSERVATIONS Competitiveness (chapter 5): the estimated level of error is 4,4 %, a decrease compared with the results of 2014 (5,6 %). Much of the expenditure is made on a cost reimbursement basis and the errors here essentially reflect different categories of ineligible cost (notably personnel costs, other direct costs, and indirect costs). THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Cohesion (chapter 6): the estimated level of error is 5,2 %, lower than the results of 2014 (5,7 %). Almost all spending in this area takes the form of cost reimbursement. Ineligible costs in expenditure declarations and ineligible projects account for three quarters of the error. Serious infringements of public procurement rules make up one seventh of the total (one half in 2014). The Commission refers to its reply to paragraphs 6.11 and 6.76. Natural Resources (chapter 7): the estimated level of error is 2,9 %, similar to 2014 (3,6 %). European Agriculture Guarantee Fund (EAGF) accounts for more than three-quarters of expenditure in this area and is less affected by error (2,2 %) than Rural Development (5,3 %), where we have long found higher levels of error for investment spending (where agricultural businesses are subsidised on the basis of reimbursement of eligible spending). Nearly two thirds of MFF heading2 expenditure is area related direct support to farmers. Consequently, overstated claims of agricultural areas were the highest contributor to the estimated level of error (more than a half). Errors related to ineligible beneficiaries, activities or expenditure contributed to one fifth. Breaches of public procurement rules increased on the level found in 2014. The lower error rate for EAGF reflects the effectiveness of the IACS, notably the LPIS, where adequately implemented, in preventing errors from the very beginning of the process. The Commission is of the opinion that the IACS as a whole is a solid system for the management of CAP expenditure (see paragraphs 7.16 to 7.18). The higher error rate in rural development is partly due to the complexity of the rules governing some measures. However, the error rate for rural development, although above the materiality threshold has fallen in recent years. Global Europe (chapter 8): the estimated level of error is 2,8 %, close to the results in 2014 (2,7 %). Commission clearings where works, services or supplies had not been delivered together with ineligible costs reimbursed by the Commission account for two thirds of the total. Administration (chapter 9): the estimated level of error is 0,6 %, in line with the results in 2014 (0,5 %). Most spending in this area is on the salaries, pensions and allowances paid by EU institutions and agencies. A small number of errors in allowances accounted for most of the error in this area, which is, in total, not material. THE COURT’S OBSERVATIONS We do not estimate levels of error for other areas of spending, including MFF heading 3 (see chapter 8). In total, expenditure covered by our statement of assurance in these areas amounted to 3 billion euro (2,1 % of the spending covered by our audit). Work performed in these areas continues to contribute to our overall conclusions on 2015. For 2015, the estimated level of error on shared management expenditure taken as a whole amounts to 4,0 % (2014: 4,6 %) and on all other forms of operational expenditure is 3,9 % (2014: 4,6 %). The estimated level of error for administrative expenditure is 0,6 % (2014: 0,5 %). The Commission significantly increased estimates of levels of error in successive documents… Each Commission directorate-general produces an annual activity report, which includes a declaration in which the Director-General provides assurance that the report properly presents financial information and that transactions under his/her responsibility are legal and regular. They also provide an account of the achievement of the key policy objectives (discussed in chapter 3) and a management report of the Director-General to the Commissioners. In 2015 the Commission simplified the structure of the annual activity reports and gave DGs more flexibility in presenting them. ECA 2015 audit results compared with estimated amounts at risk at payment and corrective capacity in the 2015 management and performance report for the EU budget and annual activity reports Annual report chapter Estimated level of error (%) Confidence interval (%) Commission annual activity reports Directorate General Gross financial exposure on payments reported by DGs (%) Commission's adjusted financial exposure (%) (see paragraph 1.35) Commission's corrective capacity (%) Lower error limit (LEL) Upper error limit (UEL) Lowest value Highest value Lowest value Highest value Chapter 5 — Competitiveness 4,4 2,0 6,7 CONNECT, EAC, EACEA, EASME, ECFIN, ENER, ERCEA, FISMA, GROW, INEA, MOVE, REA, RTD and TAXUD 1,2 1,4 1,9 2,2 0,9 Chapter 6 — Cohesion 5,2 2,8 7,6 EMPL and REGIO 2,5 4,3 2,9 5,1 1,6-2,7 Chapter 7 — Natural resources 2,9 1,7 4,2 AGRI, CLIMA, ENV, MARE and SANTE 2,0 2,0 2,0 2,0 1,9 Global Europe and EDFs 3,1 1,7 4,5 AGRI, DEVCO, EACEA, ECFIN, ECHO, EMPL, FPI, HOME, NEAR, REGIO, TRADE 2,2 2,3 2,6 2,7 0,6 Chapter 9 — Administration 0,6 0,0 1,2 Administration 0,3 0,4 0,3 0,4 0,0 Source: European Court of Auditors. Source: Commission annual activity reports. Source: European Court of Auditors based on data provided by DG Budget. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The Commission reports three sets of estimated levels of error: (21) the original estimates reported to the Commission by bodies and authorities set up under sectoral legislation and other entities (made available to us in April and May); (22) estimates (which are in some cases significantly higher) in the annual activity reports adjusted on the basis of the Commission’s assessment of available information, also available to us in April and May; (23) adjusted figures published in July 2016 in the annual management and performance report for the EU budget (AMPR), which take account of differences between cash flows and clearings. Txt. 10. Our estimated level of error is based on the examination of a statistical sample of interim and final payments and clearings (see Figure 1.8). Unlike US bodies managing federal funds (see Annex 1.2) the European Commission is not required to produce its own, statistically-valid, estimate of the level of error. The Commission decided to adjust the estimates for the AMPR in order to be more comparable to the Court’s estimated level of error. The Commission considers that reference to the statistical methodology used by US bodies managing federal funds is not appropriate for funds under shared management. Under shared management, the Commission relies not on audits of expenditure made by its own services at EU level but on audit bodies in sovereign Member States auditing expenses made by these Member States, frequently with national co-financing. Therefore the Commission considers that the production of an EU-wide statistically valid estimate of the level of error by itself auditing beneficiaries would not provide added value for shared management expenditure. In line with the requirements of the legal framework, the Commission will continue to cooperate with Member States’ audit bodies and define in common a statistically valid approach to estimate the level of error at programme level, and to use such error rates after validation. The Commission thus implements the single audit principle by focusing its activity on auditing the Member State auditors and validating their audit work. This is based upon the Court’s opinion No 2/2004 on the single audit model and a proposal for a Community internal control framework. This approach is fully in line with the objective of reducing the administrative burden on beneficiaries and reduces also the need of audit posts in the EC. The Financial Regulation requires us to transmit our preliminary observations to the Commission on the same date that the Commission is required to sign a synthesis report. This year, the Commission forwarded us its AMPR, including the synthesis report on 2015 on 5 July 2016, the date of our adversarial meeting. We therefore do not offer substantive comments on the synthesis report in this report (which this year forms a part of the AMPR). The synthesis report and the AMPR do not form a part of the financial statements and we do not offer any assurance on its contents. … and calculation of corrective capacity remains a challenge for the Commission Since 2014 DGs present their estimate of the level of error alongside their estimate of future corrective capacity. The Commission defines this as a conservative estimate of the amount of expenditure in breach of applicable regulatory and contractual provisions, that the DG will identify and correct through checks made after the payment is authorised. For the calculation of the corrective capacity DG BUDG provides directors-general with the average amount of recoveries and financial corrections made since 2009 as recorded in the accounts. Of the 48 DGs and EAs, 29 have made appropriate adjustments to this figure in order to present a more conservative estimate, notably by removing ex-ante and other non-relevant corrections and recoveries and one-off events, or by opting for an alternative basis. In the AMPR the Commission's estimate for future corrections affecting 2015 spending is, on this basis, 2,1 billion euro (2014: 2,7 billion euro). The Commission also provides a higher estimate of potential future corrections of 2,7 billion euro. The Commission provides in its AMPR a range concerning the estimated future corrections which lies between 2,1 and 2,7 billion euro. Both estimates result from a conservative approach. The 2015 FSDA reports a total implemented financial corrections and recoveries amounting to 3,9 billion euro. As in previous years, the amount recorded in the accounts covered a wide range of measures, which applied both before and after the Commission made (or accepted) expenditure. In Figure 1.9 we show our classification of the corrections and recoveries and their distribution by spending area. Classification and distribution by spending area of corrective measures implemented in 2015 Source: European Court of Auditors based on FSDA of the 2015 consolidated accounts of the EU and underlying data. THE COURT’S OBSERVATIONS In 2015 around 1,2 billion euro of the corrections and recoveries reported by the Commission were at source (applied before it accepted expenditure). This represents more than a quarter of the total reported as financial corrections and recoveries. The Commission therefore did not record these amounts as expenditure: they represent deductions from requested amounts before approval. Under direct and indirect management, these deductions — normally recorded as recoveries, represented more than two-thirds of the total corrective activity reported for 2015. The remainder (2,7 billion euro) related to claims the Commission had already accepted. Of this amount: Pt_a_. 4. Around 1,1 billion euro represents withdrawals by Member States of previously accepted claims for reimbursement for cohesion projects or expenditure and replacement with new projects or expenditure. These withdrawals are presented together with new projects and expenditure and so form part of the population from which we draw our sample. They do not result in the return of funds to the EU budget: they are effective as a corrective measure where the replacement projects and expenditure are free of material error. Pt_b_. 4. Around 1,5 billion euro concerns agriculture conformity decisions that the Commission recorded as assigned revenue, available to fund agricultural spending. These corrections are often made on a flat-rate basis and relate to system weaknesses rather than errors at the level of the final beneficiaries. Txt. 11. Timely corrective action reduced our estimated level of error We seek to take account of corrective measures applied by the Member States and the Commission where these are made prior to payment or to our examination. We check the application of these corrections (which include recoveries from the beneficiaries and corrections at project level) and adjust the quantification of error whenever appropriate. However the impact of corrective measures varies significantly between different areas of spending, and between the different corrective measures. For 16 transactions sampled in 2015, the Commission and Member State authorities had applied corrective measures that directly affected the transaction concerned and were relevant for our calculations. These corrective measures reduced our estimated level of error by 0,5 percentage points (2014: 1,1 percentage points). Changes in the number of transactions affected and in the impact on our estimated level of error do not indicate that corrective action has become more or less effective as these measures apply to a relatively small share of our sample and fluctuation from year to year is to be expected. WE SEND CASES OF SUSPECTED FRAUD TO OLAF We send cases of suspected fraud identified in our audit to the European Anti-Fraud Office (OLAF) for analysis and possible investigation. We cannot comment on individual cases or on OLAF’s response to these. However, we note that during the 2015 audit: (26) we assessed the legality and regularity of some 1200 transactions; (27) we found 12 instances of suspected fraud (2014: 22) which we forwarded to OLAF; (28) the most frequent instances of suspected fraud concerned conflicts of interest and the artificial creation of conditions to receive subsidy followed by declarations of costs not meeting the eligibility criteria. Txt. 12. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES CONCLUSIONS The key function of this chapter is to support the audit opinion presented in the statement of assurance. Audit results Our audit results, while showing improvements in some areas, are broadly consistent with those of previous years. Our results as set out in Figures 1.5 and 1.6 show a strong relationship between the basis for payment and levels of error. We continue to find that reimbursement spending is affected by much higher levels of error than spending on an entitlement basis. The Commission has taken further steps to improve its quantification of amounts at risk and corrective capacity. However, there is scope for improvement in the Commission’s assessment of both these issues. The Commission has improved its methodology and adjusted the figures in the 2015 Annual Management and Performance report for the EU budget, where necessary, to ensure that the figures reported provide a conservative estimate of the amounts at risk and the corrective capacity. Taking into account the new legal framework for 2014-2020, future adjustments to the methodology may be appropriate.

THE COURT'S STATEMENT OF ASSURANCE PROVIDED TO THE EUROPEAN PARLIAMENT AND THE COUNCIL — INDEPENDENT AUDITOR’S REPORT We have audited:

(a) the consolidated accounts of the European Union which comprise the consolidated financial statements and the aggregated reports on the implementation of the budget for the financial year ended 31 December 2015, approved by the Commission on 11 July 2016, and

(b) the legality and regularity of the transactions underlying those accounts, as required by Article 287 of the Treaty on the Functioning of the European Union (TFEU).

Management's responsibility In accordance with Articles 317 to 318 of the TFEU and the Financial Regulation, the Commission is responsible for the preparation and fair presentation of the consolidated accounts of the European Union on the basis of internationally accepted accounting standards for the public sector and for the legality and regularity of the transactions underlying them. This responsibility includes the design, implementation and maintenance of internal control relevant to the preparation and presentation of financial statements that are free from material misstatement, whether due to fraud or error. The Commission bears the ultimate responsibility for the legality and regularity of the transactions underlying the accounts of the European Union (Article 317 of the TFEU). Auditor's responsibility Our responsibility is to provide, on the basis of our audit, the European Parliament and the Council with a statement of assurance as to the reliability of the accounts and the legality and regularity of the transactions underlying them. We conducted our audit in accordance with the IFAC International Standards on Auditing and Codes of Ethics and the INTOSAI International Standards of Supreme Audit Institutions. These standards require that we comply with ethical requirements and plan and perform the audit to obtain reasonable assurance as to whether the consolidated accounts of the European Union are free from material misstatement and the transactions underlying them are legal and regular. An audit involves performing procedures to obtain audit evidence about the amounts and disclosures in the consolidated accounts and the legality and the regularity of the transactions underlying them. The procedures selected depend on the auditor's judgment, including an assessment of the risks of material misstatement of the consolidated accounts and of material non-compliance of the underlying transactions with the requirements of the legal framework of the European Union, whether due to fraud or error. In making those risk assessments, internal control relevant to the preparation and fair presentation of the consolidated accounts and legality and regularity of underlying transactions, is considered in order to design audit procedures that are appropriate in the circumstances, but not for the purpose of expressing an opinion on the effectiveness of internal control. An audit also includes evaluating the appropriateness of accounting policies used and the reasonableness of accounting estimates made, as well as evaluating the overall presentation of the consolidated accounts. For revenue, our examination of value added tax and gross national income-based own resources takes as its starting point the relevant macroeconomic aggregates on which these are calculated, and assesses the Commission's systems for processing these until the contributions of the Member States have been received and recorded in the consolidated accounts. For traditional own resources, we examine the accounts of the customs authorities and analyse the flow of duties until the amounts are received by the Commission and recorded in the accounts. For expenditure, we examine payment transactions when expenditure has been incurred, recorded and accepted. This examination covers all categories of payments (including those made for the purchase of assets) other than advances at the point they are made. Advance payments are examined when the recipient of funds is required to provide or provides justification for their proper use and the advance payment is cleared or becomes recoverable. We consider that the audit evidence obtained is sufficient and appropriate to provide a basis for our opinions. Reliability of the accounts Opinion on the reliability of the accounts In our opinion, the consolidated accounts of the European Union for the year ended 31 December 2015 present fairly, in all material respects, the financial position of the Union as at 31 December 2015, the results of its operations, its cash flows, and the changes in net assets for the year then ended, in accordance with the Financial Regulation and with accounting rules based on internationally accepted accounting standards for the public sector. Legality and regularity of the transactions underlying the accounts Revenue Opinion on the legality and regularity of revenue underlying the accounts In our opinion, revenue underlying the accounts for the year ended 31 December 2015 is legal and regular in all material respects. Payments Basis for adverse opinion on the legality and regularity of payments underlying the accounts Expenditure recorded in 2015 under the multi-annual financial framework headings covering operational spending is materially affected by error. Our estimated level of error for payments underlying the accounts is 3,8 %. Our overall conclusion is driven by the higher estimated level of error for spending on a reimbursement basis and is corroborated by the Commission’s analysis of amounts at risk presented in the annual management and performance report for the EU budget. Adverse opinion on the legality and regularity of payments underlying the accounts In our opinion, because of the significance of the matters described in the basis for adverse opinion on the legality and regularity of payments underlying the accounts paragraph, the payments underlying the accounts for the year ended 31 December 2015 are materially affected by error. Other information The Financial Statement Discussion and Analysis is not a part of the financial statements. The information given in the Financial Statement Discussion and Analysis is consistent with the financial statements. 14 July 2016 Vítor Manuel da SILVA CALDEIRA President European Court of Auditors 12, rue Alcide De Gasperi, 1615 Luxembourg, LUXEMBOURG THE COURT’S OBSERVATIONS INTRODUCTION EU spending is a significant tool for achieving policy objectives EU spending amounted to 145,2 billion euro in 2015. It represents a significant instrument for achieving policy objectives. Other significant instruments are the use of legislation and the freedom for goods, services, capital and people to move throughout the Union. EU spending represents 2,1 % of EU Member States’ (see Figure 1.1) total general government expenditure, and 1,0 % of EU gross national income. As set out in the Treaty, we:

check whether the EU budget is used in accordance with applicable laws and regulations;

give our opinion on the accounts of the EU;

report on whether EU spending is economic, efficient and effective; and

advise on proposed legislation with a financial impact.

The work we do for the statement of assurance fulfils the first and second of these objectives. It helps us identify risks affecting the economy, efficiency and effectiveness of spending. It provides a key input to our opinions on proposed financial legislation. 2015 EU spending in each Member State as a share of gross national income (GNI) and general government expenditure Member States GNI: Agreed set of forecasts of traditional own resources and VAT/GNI bases — 19.5.2016 — European Commission. Member States — general government expenditure — Eurostat — annual national accounts. See Annex 1.3 for explanation of country codes. Covers the expenditure, salaries and other costs of EU bodies located in the Member States. Source: European Union spending (European Commission accounting data, compiled by ECA). THE COURT’S OBSERVATIONS This chapter of the annual report:

sets out the background to our statement of assurance and summarises our audit approach (see Annex 1.1), findings and conclusions;

discusses key audit matters arising from the audit of the Union’s financial statements;

compares our results on the regularity of spending in 2015 with findings in recent years and with the estimates of the European Commission.

Annex 1.2 discusses United States (US) and EU approaches to testing and reporting on improper and irregular payments.

THE ACCOUNTS WERE NOT AFFECTED BY MATERIAL MISSTATEMENTS Our observations concern the consolidated accounts (the accounts) of the European Union for the financial year 2015. We received them, together with the accounting officer’s letter of representation, on 12 July 2016. The accounts are accompanied by a Financial Statement Discussion and Analysis. This analysis is not covered by our audit opinion. In accordance with the auditing standards, we have, however, assessed its consistency with information in the accounts. The accounts published by the Commission show that, at 31 December 2015, total liabilities amounted to 226,1 billion euro compared to 153,7 billion euro of total assets. The deficit for 2015 amounts to 13,0 billion euro. Our audit found that the accounts were not affected by material misstatements. We present our observations on the financial and budgetary management of EU funds in Chapter 2. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES In 2015 the Commission further strengthened alignment with international standards and with the Recommended Practice Guideline for preparing and presenting Financial Statement Discussion and Analysis. KEY AUDIT MATTERS Key audit matters are those matters that, in our professional judgment, were of most significance in our audit of the financial statements of the current period. These matters were addressed in the context of our audit of the financial statements as a whole, and in forming our opinion thereon, but we do not provide a separate opinion on these matters. We sought additional information from the Commission to support the valuation of financial instruments under shared management Authorities in the Member States transfer a part of the funding advanced by the Commission into financial instruments in the form of loans, equity instruments or guarantees. For the 2007-2013 MFF, EU law did not require these authorities to produce periodic reports on sums held in these instruments in time for preparation of the accounts. The Commission therefore estimated the use made of these advances, assuming that funds would be used in full and evenly over the period of operation (initially up to 31 December 2015 but later extended — without legal authority — to 31 March 2017). By the end of 2014, of the 16 billion euro (including national contributions) transferred into cohesion financial instruments under shared management, only 9,2 billion euro (57 %) had been passed to projects. The Commission recalls its disagreement with the Court’s statement concerning the extension of the period of operation for financial instruments under shared management (see also the Commission’s replies to paragraphs 6.41 and 6.45). THE COURT’S OBSERVATIONS We required the Commission to confirm the validity of its methodology for estimating the amount presented in its balance sheet (3,3 billion euro at 31 December 2015). The Commission therefore launched a data collection exercise, targeted at the Member States presenting the highest balances. We reviewed the procedure put in place for collecting the data from the Member States concerned and taking into consideration the results obtained, we conclude that the overall amount of pre-financing related to financial instruments presented in the balance sheet is fairly stated. We assessed the accounting treatment of the European Fund for Strategic Investments (EFSI) guarantee operations In 2015 the European Fund for Strategic Investments (EFSI) was launched. EFSI is a joint initiative launched by the EU and the EIB to encourage investment across the EU, by increasing the risk bearing capacity of the EIB Group. Under this initiative, the EU budget will provide a guarantee of up to 16 billion euro to cover potential losses on EFSI operations (12,3 billion euro for a debt portfolio and 3,7 billion euro for an equity portfolio) as explained in paragraphs 2.32-2.38. We focused on this area because the set-up of EFSI has given rise to a number of complex accounting issues resulting from the arrangements between the EU and the EIB. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Under the EFSI debt portfolio, the EU budget guarantee covers the first loss of a portfolio of financing operations entered into by the EIB. The Commission therefore accounts for it as a financial guarantee liability. Under the EFSI equity portfolio, the EIB makes equal, parallel investments at its own risk and at the risk of the EU. For the equity investments made by the EIB at the risk of the EU, the EU budget bears the whole risk and each year end the Commission transfers to the EIB a sum equal to the negative value adjustments (unrealised losses), realised losses at disinvestment and EIB funding costs, net of receipts from interests, dividends and realised gains. There is no transfer from the EIB to the EU for unrealised gains above the initial value of the investment. Equity portfolio investments (concerning only the Infrastructure and Innovation Window) covered by the EU budget guarantee amounted to 7,6 million euro at year end. The Accounting Officer exercised significant judgement when assessing and concluding that the EFSI did not meet the criteria for classification as a joint operation. Other complex accounting issues relate to the recognition and measurement of the financial guarantee liability for the debt portfolio, the EU guarantee of equity investments and the recognition of potential gains under the equity portfolio. The Accounting Officer of the Commission exercised, as he is required to do frequently for many other accounting issues, his professional judgement when taking a decision on the accounting treatment for EFSI. This judgement was based on an in-depth analysis of the facts by the accounting services of the Commission and the unanimous support of the EU advisory group on accounting standards. We focused audit work on analysing the control indicators. Had the Commission treated EFSI as a joint operation, this would require some additional disclosures on ownership interests and voting rights, inclusion in note 9 of the financial statements (scope of consolidation), and recognition of an EU share of the assets, liabilities, costs and revenues of the equity portfolio. However, at the end of 2015, this would have had no impact on the net position of the EU. The Commission has included in the notes to the 2015 accounts all mandatory disclosures required by the EU accounting rules in force for these accounts. In addition to this, both the notes and the FSDA provide extensive supplementary information on the nature, extent and financial effects of the EFSI for the EU. For the debt portfolio, we also reviewed the adequacy of the accounting policies and found the accounting treatment to be consistent and in line with international standards. For the equity portfolio we consider that if future years show an increase in the value of equity investments above their initial valuation this should be recognised in the EU accounts. At the end of 2015, the sums involved (see paragraph 1.13) do not have a material impact on the EU consolidated accounts. In the future, the impact of any gains on the net position of the EU would be the same whether the Commission accounts for these operations as a joint operation or a derivative. While we will keep this issue under review, we have concluded that it does not have an impact on our audit opinion for 2015. The Commission welcomes that the Court finds the accounting treatment of the debt portfolio to be consistent and in line with international standards. Discussions on the equity portfolio, which was not implemented in 2015, are still ongoing between the Commission and the EIB Group. The accounting treatment of the equity operations will be established by the Accounting Officer after consulting the EU Advisory Group of Experts on Accounting Standards once the amended legal basis has been finalised. We assessed the increase in accrued charges presented in the accounts In the final accounts of 2015, the Commission estimated that incurred eligible expenses due to the beneficiaries but not yet reported to it amounted to 106 billion euro. It recorded these as accrued expenses. This represents an increase of 18 billion euro on 2014. The Commission will use the experience with the estimation of incurred eligible expenses to further improve the cut-off exercise in the future. We examined the methodology and control systems applied in the most significant directorates-general to the year-end estimates and the cost statements submitted by Member States in the first quarter of 2016. We drew samples on invoices and pre-financings and carried out work on these elements to address the risk that the accrual was misstated. We sought additional explanation from the accounting officer on the increase in the accrued charges as presented in the accounts. The work we performed on the provisional accounts and the net adjustments of 3,3 billion euro made in the final accounts enable us to conclude that the overall amount of accrued charges in the final consolidated balance sheet is fairly stated. THE COURT’S OBSERVATIONS REGULARITY OF TRANSACTIONS: ALTHOUGH OUR ESTIMATED LEVEL OF ERROR REMAINS MATERIAL, THERE WERE SOME AREAS OF IMPROVEMENT We examine EU revenue and spending to assess whether it is in accordance with applicable laws and regulations. We present our audit results on revenue in chapter 4 and on expenditure in chapters 5 to 9 (see Figure 1.2).

(a) Revenue was free from material error (see paragraph 4.22).

(b) In expenditure, we continue to find a material level of error. The estimated level of error in expenditure was 3,8 %, a lower level than in 2014 (4,4 %) (see Figure 1.2). (i) Spending on MFF heading 5 (Administration) was free from material error (see paragraph 9.16). (ii) Operational spending was affected by material error (see paragraphs 5.35 to 5.36, 6.76 to 6.77, 7.64 to 7.65 and 8.33 to 8.34) with much higher levels of error affecting expenditure driven by the reimbursement of costs (see paragraph 1.22).

(b)

(i) Spending on MFF heading 5 (Administration) was free from material error (see paragraph 9.16).

(ii) Operational spending was affected by material error (see paragraphs 5.35 to 5.36, 6.76 to 6.77, 7.64 to 7.65 and 8.33 to 8.34) with much higher levels of error affecting expenditure driven by the reimbursement of costs (see paragraph 1.22).

2015 Summary of audit results on regularity of transactions Source: European Court of Auditors. Annual report chapter MFF headings Transactions subject to audit (billion euro) Estimated level of error 2015 (%) Confidence interval (%) Estimated level of error 2014 (%) Lower error limit (LEL) Upper error limit (UEL) Competitiveness Heading 1.a 14,5 4,4 2,0 6,7 5,6 Cohesion Heading 1.b 53,9 5,2 2,8 7,6 5,7 Natural Resources Heading 2 58,6 2,9 1,7 4,2 3,6 Global Europe Heading 4 6,9 2,8 1,0 4,6 2,7 Administration Heading 5 9,0 0,6 0,0 1,2 0,5 Other Heading 3, 6 and other 3,0 — — — — Total 145,9 3,8 2,7 4,8 4,4 Revenue 153,8 0 0 0 0 THE COURT’S OBSERVATIONS Our 2015 audit results show a broadly similar picture to recent years Our overall estimated level of error, although improved on recent years, continues to exceed significantly our benchmark for materiality (see Figure 1.3). The estimated level of error (2013 to 2015) Source: European Court of Auditors (see Figure 1.2). THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Reimbursement of costs is linked to higher levels of error EU spending is dominated by two types of expenditure programmes. The differences between two bases for payment (cost reimbursement and entitlement) explain a large part of the variation in the estimated level of error. 1.21.

Cost reimbursement; where the EU reimburses eligible costs for eligible activities. They include research projects (chapter 5), training schemes (chapter 6), as well as projects in regional and rural development (chapters 6 and 7) and development projects (chapter 8). In these schemes, recipients must provide information supporting their claim to be engaged in an activity eligible for support and to have incurred costs for which they are entitled to reimbursement. The requirement to provide accurate information gives rise to complicated eligibility rules (what can be claimed) and formal requirements (how to incur costs, e.g. public procurement or state aid rules).

The Commission considers that the impact of the complicated eligibility rules referred to by the Court is significantly mitigated when simplified cost options are used.

Entitlement; where payment is based on meeting certain conditions. Such programmes include student and research fellowships (chapter 5), budget support (chapter 8), salaries and pensions (chapter 9). Direct aid for farmers (chapter 7) is based on entitlement and calculated on the basis of information confirmed or declared by farmers; the risk linked to potentially incorrect declarations is largely mitigated by the simpler nature of the information expected from beneficiaries, which can, in large part, be verified before payment through, for example, comparisons with databases and orthophotos.

THE COURT’S OBSERVATIONS Eligibility errors in cost reimbursement schemes dominate the errors detected for 2015.

(a) For reimbursement expenditure the estimated level of error is 5,2 % (2014: 5,5 %). Typical errors in this area include: (1) ineligible costs included in cost claims; (2) ineligible projects, activities and beneficiaries; and (3) serious infringements of public procurement rules (Figure 1.4).

ineligible costs included in cost claims;

ineligible projects, activities and beneficiaries; and

serious infringements of public procurement rules (Figure 1.4).

(b) For entitlement expenditure, the estimated level of error is 1,9 % (2014: 2,7 %). Typical errors in this area include small over-declarations by farmers of farm and field areas (paragraph 7.16).

Breakdown of overall estimated level of error by type The 2014 comparator for Errors by Commission and intermediary bodies is the total of Administrative errors in natural resources and Other error types. Source: European Court of Auditors. THE COURT’S OBSERVATIONS Figure 1.5 is based on our transaction testing of EU spending over the last two years for the annual reports on the general budget and the European Development Funds. In general, the more complex the information that is required from beneficiaries, the higher the risk of error. The relationship between basis for payment and estimated level of error in EU transactions (2014-2015) Reimbursement for Global Europe includes multi-donor projects which in practice have many of the attributes of entitlement spending and are affected by lower levels of error. Source: European Court of Auditors. THE COURT’S OBSERVATIONS We analysed other risk factors and their relationship with the estimated level of error. We found that within direct management spending two risk factors have a significant impact on the level of error.

In areas where the rules provide for an audit or inspection before the payment is made we find, in general, a lower level of error.

Reimbursements to projects involving a number of partners (e.g. consortia) have higher levels of error than reimbursements to a single beneficiary.

Management mode has a limited impact on levels of error The different risk patterns of reimbursement and entitlement schemes, rather than management modes, influences the level of errors in different areas of spending.The estimated level of error shows a much stronger relationship with the basis for payment than with the management mode. We found the highest levels of error over the last two years in Competitiveness (managed directly by the Commission and indirectly through entrusted entities) and Cohesion (under shared management). Reimbursement schemes dominate spending in both of these areas (see Figure 1.6). Average estimated level of error by management mode and by basis for payment (2014-2015) (1) The two sides to the graph present the same amount of expenditure. On the left hand side spending is split on the basis of management mode, showing that the great majority of operational spending is under shared management. The right hand side shows spending split by the basis for payment and shows that reimbursement spending is significantly greater than entitlement spending. The vertical position of the bars indicates the estimated level of error for each. This shows the limited impact of management mode and the significant impact of basis for payment. Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Results in different areas of spending show distinct patterns of error As shown in Figure 1.7, Cohesion was the biggest contributor to our estimated level of error for 2015, followed by Natural Resources, Competitiveness and Global Europe. This distribution is in line with our findings for 2014. The Commission shares the Court’s assessment of the errors and will duly follow them up to protect the EU budget. Contribution of different MFF headings to the overall estimated level of error (2015) Source: European Court of Auditors. THE COURT’S OBSERVATIONS Competitiveness (chapter 5): the estimated level of error is 4,4 %, a decrease compared with the results of 2014 (5,6 %). Much of the expenditure is made on a cost reimbursement basis and the errors here essentially reflect different categories of ineligible cost (notably personnel costs, other direct costs, and indirect costs). THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Cohesion (chapter 6): the estimated level of error is 5,2 %, lower than the results of 2014 (5,7 %). Almost all spending in this area takes the form of cost reimbursement. Ineligible costs in expenditure declarations and ineligible projects account for three quarters of the error. Serious infringements of public procurement rules make up one seventh of the total (one half in 2014). The Commission refers to its reply to paragraphs 6.11 and 6.76. Natural Resources (chapter 7): the estimated level of error is 2,9 %, similar to 2014 (3,6 %). European Agriculture Guarantee Fund (EAGF) accounts for more than three-quarters of expenditure in this area and is less affected by error (2,2 %) than Rural Development (5,3 %), where we have long found higher levels of error for investment spending (where agricultural businesses are subsidised on the basis of reimbursement of eligible spending). Nearly two thirds of MFF heading2 expenditure is area related direct support to farmers. Consequently, overstated claims of agricultural areas were the highest contributor to the estimated level of error (more than a half). Errors related to ineligible beneficiaries, activities or expenditure contributed to one fifth. Breaches of public procurement rules increased on the level found in 2014. The lower error rate for EAGF reflects the effectiveness of the IACS, notably the LPIS, where adequately implemented, in preventing errors from the very beginning of the process. The Commission is of the opinion that the IACS as a whole is a solid system for the management of CAP expenditure (see paragraphs 7.16 to 7.18). The higher error rate in rural development is partly due to the complexity of the rules governing some measures. However, the error rate for rural development, although above the materiality threshold has fallen in recent years. Global Europe (chapter 8): the estimated level of error is 2,8 %, close to the results in 2014 (2,7 %). Commission clearings where works, services or supplies had not been delivered together with ineligible costs reimbursed by the Commission account for two thirds of the total. Administration (chapter 9): the estimated level of error is 0,6 %, in line with the results in 2014 (0,5 %). Most spending in this area is on the salaries, pensions and allowances paid by EU institutions and agencies. A small number of errors in allowances accounted for most of the error in this area, which is, in total, not material. THE COURT’S OBSERVATIONS We do not estimate levels of error for other areas of spending, including MFF heading 3 (see chapter 8). In total, expenditure covered by our statement of assurance in these areas amounted to 3 billion euro (2,1 % of the spending covered by our audit). Work performed in these areas continues to contribute to our overall conclusions on 2015. For 2015, the estimated level of error on shared management expenditure taken as a whole amounts to 4,0 % (2014: 4,6 %) and on all other forms of operational expenditure is 3,9 % (2014: 4,6 %). The estimated level of error for administrative expenditure is 0,6 % (2014: 0,5 %). The Commission significantly increased estimates of levels of error in successive documents… Each Commission directorate-general produces an annual activity report, which includes a declaration in which the Director-General provides assurance that the report properly presents financial information and that transactions under his/her responsibility are legal and regular. They also provide an account of the achievement of the key policy objectives (discussed in chapter 3) and a management report of the Director-General to the Commissioners. In 2015 the Commission simplified the structure of the annual activity reports and gave DGs more flexibility in presenting them. ECA 2015 audit results compared with estimated amounts at risk at payment and corrective capacity in the 2015 management and performance report for the EU budget and annual activity reports Annual report chapter Estimated level of error (%) Confidence interval (%) Commission annual activity reports Directorate General Gross financial exposure on payments reported by DGs (%) Commission's adjusted financial exposure (%) (see paragraph 1.35) Commission's corrective capacity (%) Lower error limit (LEL) Upper error limit (UEL) Lowest value Highest value Lowest value Highest value Chapter 5 — Competitiveness 4,4 2,0 6,7 CONNECT, EAC, EACEA, EASME, ECFIN, ENER, ERCEA, FISMA, GROW, INEA, MOVE, REA, RTD and TAXUD 1,2 1,4 1,9 2,2 0,9 Chapter 6 — Cohesion 5,2 2,8 7,6 EMPL and REGIO 2,5 4,3 2,9 5,1 1,6-2,7 Chapter 7 — Natural resources 2,9 1,7 4,2 AGRI, CLIMA, ENV, MARE and SANTE 2,0 2,0 2,0 2,0 1,9 Global Europe and EDFs 3,1 1,7 4,5 AGRI, DEVCO, EACEA, ECFIN, ECHO, EMPL, FPI, HOME, NEAR, REGIO, TRADE 2,2 2,3 2,6 2,7 0,6 Chapter 9 — Administration 0,6 0,0 1,2 Administration 0,3 0,4 0,3 0,4 0,0 Source: European Court of Auditors. Source: Commission annual activity reports. Source: European Court of Auditors based on data provided by DG Budget. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The Commission reports three sets of estimated levels of error:

the original estimates reported to the Commission by bodies and authorities set up under sectoral legislation and other entities (made available to us in April and May);

estimates (which are in some cases significantly higher) in the annual activity reports adjusted on the basis of the Commission’s assessment of available information, also available to us in April and May;

adjusted figures published in July 2016 in the annual management and performance report for the EU budget (AMPR), which take account of differences between cash flows and clearings.

Our estimated level of error is based on the examination of a statistical sample of interim and final payments and clearings (see Figure 1.8). Unlike US bodies managing federal funds (see Annex 1.2) the European Commission is not required to produce its own, statistically-valid, estimate of the level of error. The Commission decided to adjust the estimates for the AMPR in order to be more comparable to the Court’s estimated level of error. The Commission considers that reference to the statistical methodology used by US bodies managing federal funds is not appropriate for funds under shared management. Under shared management, the Commission relies not on audits of expenditure made by its own services at EU level but on audit bodies in sovereign Member States auditing expenses made by these Member States, frequently with national co-financing. Therefore the Commission considers that the production of an EU-wide statistically valid estimate of the level of error by itself auditing beneficiaries would not provide added value for shared management expenditure. In line with the requirements of the legal framework, the Commission will continue to cooperate with Member States’ audit bodies and define in common a statistically valid approach to estimate the level of error at programme level, and to use such error rates after validation. The Commission thus implements the single audit principle by focusing its activity on auditing the Member State auditors and validating their audit work. This is based upon the Court’s opinion No 2/2004 on the single audit model and a proposal for a Community internal control framework. This approach is fully in line with the objective of reducing the administrative burden on beneficiaries and reduces also the need of audit posts in the EC. The Financial Regulation requires us to transmit our preliminary observations to the Commission on the same date that the Commission is required to sign a synthesis report. This year, the Commission forwarded us its AMPR, including the synthesis report on 2015 on 5 July 2016, the date of our adversarial meeting. We therefore do not offer substantive comments on the synthesis report in this report (which this year forms a part of the AMPR). The synthesis report and the AMPR do not form a part of the financial statements and we do not offer any assurance on its contents. … and calculation of corrective capacity remains a challenge for the Commission Since 2014 DGs present their estimate of the level of error alongside their estimate of future corrective capacity. The Commission defines this as a conservative estimate of the amount of expenditure in breach of applicable regulatory and contractual provisions, that the DG will identify and correct through checks made after the payment is authorised. For the calculation of the corrective capacity DG BUDG provides directors-general with the average amount of recoveries and financial corrections made since 2009 as recorded in the accounts. Of the 48 DGs and EAs, 29 have made appropriate adjustments to this figure in order to present a more conservative estimate, notably by removing ex-ante and other non-relevant corrections and recoveries and one-off events, or by opting for an alternative basis. In the AMPR the Commission's estimate for future corrections affecting 2015 spending is, on this basis, 2,1 billion euro (2014: 2,7 billion euro). The Commission also provides a higher estimate of potential future corrections of 2,7 billion euro. The Commission provides in its AMPR a range concerning the estimated future corrections which lies between 2,1 and 2,7 billion euro. Both estimates result from a conservative approach. The 2015 FSDA reports a total implemented financial corrections and recoveries amounting to 3,9 billion euro. As in previous years, the amount recorded in the accounts covered a wide range of measures, which applied both before and after the Commission made (or accepted) expenditure. In Figure 1.9 we show our classification of the corrections and recoveries and their distribution by spending area. Classification and distribution by spending area of corrective measures implemented in 2015 Source: European Court of Auditors based on FSDA of the 2015 consolidated accounts of the EU and underlying data. THE COURT’S OBSERVATIONS In 2015 around 1,2 billion euro of the corrections and recoveries reported by the Commission were at source (applied before it accepted expenditure). This represents more than a quarter of the total reported as financial corrections and recoveries. The Commission therefore did not record these amounts as expenditure: they represent deductions from requested amounts before approval. Under direct and indirect management, these deductions — normally recorded as recoveries, represented more than two-thirds of the total corrective activity reported for 2015. The remainder (2,7 billion euro) related to claims the Commission had already accepted. Of this amount:

(a) Around 1,1 billion euro represents withdrawals by Member States of previously accepted claims for reimbursement for cohesion projects or expenditure and replacement with new projects or expenditure. These withdrawals are presented together with new projects and expenditure and so form part of the population from which we draw our sample. They do not result in the return of funds to the EU budget: they are effective as a corrective measure where the replacement projects and expenditure are free of material error.

(b) Around 1,5 billion euro concerns agriculture conformity decisions that the Commission recorded as assigned revenue, available to fund agricultural spending. These corrections are often made on a flat-rate basis and relate to system weaknesses rather than errors at the level of the final beneficiaries.

Timely corrective action reduced our estimated level of error We seek to take account of corrective measures applied by the Member States and the Commission where these are made prior to payment or to our examination. We check the application of these corrections (which include recoveries from the beneficiaries and corrections at project level) and adjust the quantification of error whenever appropriate. However the impact of corrective measures varies significantly between different areas of spending, and between the different corrective measures. For 16 transactions sampled in 2015, the Commission and Member State authorities had applied corrective measures that directly affected the transaction concerned and were relevant for our calculations. These corrective measures reduced our estimated level of error by 0,5 percentage points (2014: 1,1 percentage points). Changes in the number of transactions affected and in the impact on our estimated level of error do not indicate that corrective action has become more or less effective as these measures apply to a relatively small share of our sample and fluctuation from year to year is to be expected. WE SEND CASES OF SUSPECTED FRAUD TO OLAF We send cases of suspected fraud identified in our audit to the European Anti-Fraud Office (OLAF) for analysis and possible investigation. We cannot comment on individual cases or on OLAF’s response to these. However, we note that during the 2015 audit:

we assessed the legality and regularity of some 1200 transactions;

we found 12 instances of suspected fraud (2014: 22) which we forwarded to OLAF;

the most frequent instances of suspected fraud concerned conflicts of interest and the artificial creation of conditions to receive subsidy followed by declarations of costs not meeting the eligibility criteria.

THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES CONCLUSIONS The key function of this chapter is to support the audit opinion presented in the statement of assurance. Audit results Our audit results, while showing improvements in some areas, are broadly consistent with those of previous years. Our results as set out in Figures 1.5 and 1.6 show a strong relationship between the basis for payment and levels of error. We continue to find that reimbursement spending is affected by much higher levels of error than spending on an entitlement basis. The Commission has taken further steps to improve its quantification of amounts at risk and corrective capacity. However, there is scope for improvement in the Commission’s assessment of both these issues. The Commission has improved its methodology and adjusted the figures in the 2015 Annual Management and Performance report for the EU budget, where necessary, to ensure that the figures reported provide a conservative estimate of the amounts at risk and the corrective capacity. Taking into account the new legal framework for 2014-2020, future adjustments to the methodology may be appropriate.

ANNEX

Grp. 1. Our audit approach is set out in the Financial and Compliance Audit Manual available on our website. We use an assurance model to plan our work. In our planning, we consider the risk of errors occurring (inherent risk) and the risk that errors are not prevented or detected and corrected (control risk). (a) the consolidated financial statements; and (b) the aggregated reports on the implementation of the budget. (3) the financial position of the European Union at year end; (4) the results of its operations and cash flows; and (5) the changes in net assets for the year ended. Pt_a_. 2. An evaluation of the accounting control environment. Pt_b_. 2. Checks on the functioning of key accounting procedures and the year-end closure process. (c) Analytical checks (consistency and reasonableness) on the main accounting data. (d) Analyses and reconciliations of accounts and/or balances. (e) Substantive tests of commitments, payments and specific balance sheet items based on representative samples. (f) Where possible, and in accordance with international standards on auditing, the use of the work of other auditors. This is particularly the case for the audit of those borrowing and lending activities managed by the Commission for which external audit certificates are available. Grp_1_audit_approach_and_methodology_for_the_reliability_of_accounts. 2. The consolidated accounts consist of: The consolidated accounts should properly present, in all material respects: Our audit involves: (1) is incorrectly calculated; or (2) does not meet a regulatory requirement or contractual provision. (3) the final recipient of EU funds (e.g. a farmer, a research institute, a company providing publicly procured works or services) provides evidence of their proper use; and (4) the Commission (or other institution or body managing EU funds) accepts that final use of funds is justified by clearing the advance payment. Grp_we_test_transactions. 45. Within each specific assessment (chapters 5 to 9), we carry out direct tests of transactions on the basis of a representative sample of transactions. Our testing provides an estimate of the extent to which the transactions in the population concerned are irregular. Transaction testing involves examining each selected transaction. We determine whether or not the claim or payment was made for the purposes approved by the budget and specified in relevant legislation. We examine the calculation of the amount of the claim or payment (for larger claims based on a representative selection of the items on which subsidy is based). This involves tracing the transaction down from the budgetary accounts to the level of the final recipient (e.g. a farmer, the organiser of a training course, or a development aid project promoter). We test compliance at each level. There is an error when the transaction (at any level): For revenue, our examination of value added tax and gross national income-based own resources takes as a starting point the relevant macroeconomic aggregates on which these are calculated. We examine and take assurance from the Commission’s control systems for processing these until the contributions of the Member States have been received and recorded in the consolidated accounts. For traditional own resources, we examine the accounts of the customs authorities and the flow of duties until the amounts are received by the Commission and recorded in the accounts. For expenditure we examine payments when expenditure has been incurred, recorded and accepted. This examination covers all categories of payments (including those made for the purchase of assets). We do not examine advances at the point they are made. We examine advance payments when: Our audit sample is designed to provide an estimate of the level of error in the audited population as a whole. We examine larger claims or payments by selecting items (e.g. invoices of a project, parcels in a claim by a farmer, see paragraph 5) to audit within individual transactions using monetary unit sampling (MUS). Thus in case of examined items being part of a project or claim by a farmer, the error rate reported for these items does not constitute an error rate for the audited project or a claim by a farmer, but contributes to the overall evaluation of EU expenditure. We do not examine transactions in every Member State, beneficiary state and/or region each year. The examples provided in the annual report are for illustrative purposes and demonstrate the most typical errors found. The naming of certain Member States, beneficiary states and/or regions does not mean that the examples presented do not occur elsewhere. The illustrative examples presented in this report do not form a basis for conclusions to be drawn on the Member States, beneficiary states and/or regions concerned. Our approach is not designed to gather data on the frequency of error in the population. Therefore, figures presented on frequency of error are not an indication of the frequency of error in EU-funded transactions or in individual Member States. Our sampling approach applies different weighting to different transactions. Our sampling reflects the value of the expenditure concerned and the intensity of audit work. This weighting is removed in a frequency table which therefore gives as much weight to rural development as to direct support in the area of Natural Resources and to Social Fund expenditure as to regional and cohesion payments in the Cohesion chapter. The relative frequency of error in samples drawn in different Member States cannot be a guide to the relative level of error in different Member States. Grp_we_evaluate_and_present_the_results_of_transaction_testing. 11121314. Errors in transactions occur for a variety of reasons. They take a number of different forms depending on the nature of the breach and specific rule or contractual requirement not followed. Individual transactions may be wholly or partially affected by error. Errors detected and corrected before and independently of the checks carried out by us are excluded from the calculation and frequency of error. They demonstrate that the control systems have worked effectively. We consider whether individual errors are quantifiable or non-quantifiable. We take account of the extent to which it is possible to measure how much of the amount audited was affected by error. Many errors occur in the application of public procurement laws. To respect the basic principles of competition foreseen in EU laws and regulations, significant procedures must be advertised. Bids must be evaluated according to specified criteria. Contracts may not be artificially split to avoid breaching thresholds. Our criteria for quantification of public procurement errors are described in the document Non-compliance with the rules on public procurement — types of irregularities and basis for quantification. Our quantification may differ from that used by the Commission or Member States when deciding how to respond to the misapplication of the public procurement rules. Grp_level_of_error. 151617. We estimate the level of error using the most likely rate of error (MLE). We do this for each MFF heading, and for spending from the budget as whole. Only quantified errors are part of the calculation. The MLE percentage is a statistical estimate of the likely percentage of error in the population. Examples of errors are quantifiable breaches of applicable regulations, rules, and contract and grant conditions. We also estimate the lower error limit (LEL) and the upper error limit (UEL) (see illustration below). The percentage of the shaded area below the curve indicates the probability that the level of error of the population is between the LEL and the UEL. We plan our work on the basis of a materiality threshold of 2 %. We use the level of materiality as guidance for our opinion. We also take account of the nature, amount and context of errors when forming our opinion. Grp_we_examine_systems_and_report_the_results. 1819. The Commission, other EU institutions and bodies, Member States’ authorities, beneficiary countries and/or regions establish systems. They use these systems to manage the risks to the budget, including the regularity of transactions. Examining systems is particularly useful for identifying recommendations for improvement. Each MFF heading, including revenue, involves many individual systems. We select a sample of systems each year. We present the results with recommendations for improvement. Grp_we_arrive_at_our_opinions_in_the_statement_of_assurance. 20. All our work reported in chapters 4 to 9 forms the basis for our opinion on the regularity of transactions underlying the European Union's consolidated accounts. Our opinion is set out in the statement of assurance. In forming our opinion we consider whether an error is pervasive. Our work allows us to arrive at an informed opinion as to whether errors in the population exceed or fall within the materiality limits. Our best estimate of the level of error for overall spending in 2015 is 3,8 %. We have more than 95 % confidence that the level of error for the audited population is material. The estimated level of error found in different MFF headings varies as described in chapters 5 to 9. We assessed error as pervasive — extending across the majority of spending areas. Grp_fraud. 21. If we have reason to suspect that fraudulent activity has taken place, we report this to OLAF, the Union’s antifraud office. OLAF is responsible for carrying out any resulting investigations. We report several cases per year to OLAF. Grp_2_audit_approach_and_methodology_for_the_regularity_of_transactions. 3. The audit of the regularity of the transactions underlying the accounts involves direct testing of transactions (see Figure 1.2). We ascertain whether they are in line with the relevant rules and regulations. (a) an audit opinion on the consolidated accounts of the European Union for the financial year ended; and (b) audit opinions on the regularity of the revenue and payments underlying those accounts. Grp_3_link_between_the_audit_opinions_for_the_reliability_of_accounts_and_the_regularity_of_transactions. 22. We have issued: Our work and our opinions follow the IFAC’s International Standards on Auditing and Codes of Ethics and INTOSAI’s International Standards of Supreme Audit Institutions. These standards also provide for the situation where auditors issue audit opinions on the reliability of accounts and the regularity of transactions underlying those accounts, by stating that a modified opinion on the regularity of transactions does not in itself lead to a modified opinion on the reliability of accounts. The financial statements, on which we express an opinion, recognise that there is a material issue in relation to breaches of the rules governing expenses charged to the EU budget. Accordingly, we have decided that the existence of a material level of error affecting regularity is not in itself a reason to modify our separate opinion on the reliability of the accounts.

ANNEX 1.1 AUDIT APPROACH AND METHODOLOGY

1. Our audit approach is set out in the Financial and Compliance Audit Manual available on our website. We use an assurance model to plan our work. In our planning, we consider the risk of errors occurring (inherent risk) and the risk that errors are not prevented or detected and corrected (control risk).

PART 1 Audit approach and methodology for the reliability of accounts 2.

The consolidated accounts consist of: The consolidated accounts should properly present, in all material respects: Our audit involves: (a) the consolidated financial statements; and (b) the aggregated reports on the implementation of the budget. (3) the financial position of the European Union at year end; (4) the results of its operations and cash flows; and (5) the changes in net assets for the year ended. Pt_a_. 2. An evaluation of the accounting control environment. Pt_b_. 2. Checks on the functioning of key accounting procedures and the year-end closure process. (c) Analytical checks (consistency and reasonableness) on the main accounting data. (d) Analyses and reconciliations of accounts and/or balances. (e) Substantive tests of commitments, payments and specific balance sheet items based on representative samples. (f) Where possible, and in accordance with international standards on auditing, the use of the work of other auditors. This is particularly the case for the audit of those borrowing and lending activities managed by the Commission for which external audit certificates are available.

A.

the consolidated financial statements; and

B.

the aggregated reports on the implementation of the budget.

the financial position of the European Union at year end;

the results of its operations and cash flows; and

the changes in net assets for the year ended.

(a) An evaluation of the accounting control environment.

(b) Checks on the functioning of key accounting procedures and the year-end closure process.

(c) Analytical checks (consistency and reasonableness) on the main accounting data.

(d) Analyses and reconciliations of accounts and/or balances.

(e) Substantive tests of commitments, payments and specific balance sheet items based on representative samples.

(f) Where possible, and in accordance with international standards on auditing, the use of the work of other auditors. This is particularly the case for the audit of those borrowing and lending activities managed by the Commission for which external audit certificates are available.

PART 2 Audit approach and methodology for the regularity of transactions 3.

The audit of the regularity of the transactions underlying the accounts involves direct testing of transactions (see Figure 1.2). We ascertain whether they are in line with the relevant rules and regulations. (1) is incorrectly calculated; or (2) does not meet a regulatory requirement or contractual provision. (3) the final recipient of EU funds (e.g. a farmer, a research institute, a company providing publicly procured works or services) provides evidence of their proper use; and (4) the Commission (or other institution or body managing EU funds) accepts that final use of funds is justified by clearing the advance payment. Grp_we_test_transactions. 45. Within each specific assessment (chapters 5 to 9), we carry out direct tests of transactions on the basis of a representative sample of transactions. Our testing provides an estimate of the extent to which the transactions in the population concerned are irregular. Transaction testing involves examining each selected transaction. We determine whether or not the claim or payment was made for the purposes approved by the budget and specified in relevant legislation. We examine the calculation of the amount of the claim or payment (for larger claims based on a representative selection of the items on which subsidy is based). This involves tracing the transaction down from the budgetary accounts to the level of the final recipient (e.g. a farmer, the organiser of a training course, or a development aid project promoter). We test compliance at each level. There is an error when the transaction (at any level): For revenue, our examination of value added tax and gross national income-based own resources takes as a starting point the relevant macroeconomic aggregates on which these are calculated. We examine and take assurance from the Commission’s control systems for processing these until the contributions of the Member States have been received and recorded in the consolidated accounts. For traditional own resources, we examine the accounts of the customs authorities and the flow of duties until the amounts are received by the Commission and recorded in the accounts. For expenditure we examine payments when expenditure has been incurred, recorded and accepted. This examination covers all categories of payments (including those made for the purchase of assets). We do not examine advances at the point they are made. We examine advance payments when: Our audit sample is designed to provide an estimate of the level of error in the audited population as a whole. We examine larger claims or payments by selecting items (e.g. invoices of a project, parcels in a claim by a farmer, see paragraph 5) to audit within individual transactions using monetary unit sampling (MUS). Thus in case of examined items being part of a project or claim by a farmer, the error rate reported for these items does not constitute an error rate for the audited project or a claim by a farmer, but contributes to the overall evaluation of EU expenditure. We do not examine transactions in every Member State, beneficiary state and/or region each year. The examples provided in the annual report are for illustrative purposes and demonstrate the most typical errors found. The naming of certain Member States, beneficiary states and/or regions does not mean that the examples presented do not occur elsewhere. The illustrative examples presented in this report do not form a basis for conclusions to be drawn on the Member States, beneficiary states and/or regions concerned. Our approach is not designed to gather data on the frequency of error in the population. Therefore, figures presented on frequency of error are not an indication of the frequency of error in EU-funded transactions or in individual Member States. Our sampling approach applies different weighting to different transactions. Our sampling reflects the value of the expenditure concerned and the intensity of audit work. This weighting is removed in a frequency table which therefore gives as much weight to rural development as to direct support in the area of Natural Resources and to Social Fund expenditure as to regional and cohesion payments in the Cohesion chapter. The relative frequency of error in samples drawn in different Member States cannot be a guide to the relative level of error in different Member States. Grp_we_evaluate_and_present_the_results_of_transaction_testing. 11121314. Errors in transactions occur for a variety of reasons. They take a number of different forms depending on the nature of the breach and specific rule or contractual requirement not followed. Individual transactions may be wholly or partially affected by error. Errors detected and corrected before and independently of the checks carried out by us are excluded from the calculation and frequency of error. They demonstrate that the control systems have worked effectively. We consider whether individual errors are quantifiable or non-quantifiable. We take account of the extent to which it is possible to measure how much of the amount audited was affected by error. Many errors occur in the application of public procurement laws. To respect the basic principles of competition foreseen in EU laws and regulations, significant procedures must be advertised. Bids must be evaluated according to specified criteria. Contracts may not be artificially split to avoid breaching thresholds. Our criteria for quantification of public procurement errors are described in the document Non-compliance with the rules on public procurement — types of irregularities and basis for quantification. Our quantification may differ from that used by the Commission or Member States when deciding how to respond to the misapplication of the public procurement rules. Grp_level_of_error. 151617. We estimate the level of error using the most likely rate of error (MLE). We do this for each MFF heading, and for spending from the budget as whole. Only quantified errors are part of the calculation. The MLE percentage is a statistical estimate of the likely percentage of error in the population. Examples of errors are quantifiable breaches of applicable regulations, rules, and contract and grant conditions. We also estimate the lower error limit (LEL) and the upper error limit (UEL) (see illustration below). The percentage of the shaded area below the curve indicates the probability that the level of error of the population is between the LEL and the UEL. We plan our work on the basis of a materiality threshold of 2 %. We use the level of materiality as guidance for our opinion. We also take account of the nature, amount and context of errors when forming our opinion. Grp_we_examine_systems_and_report_the_results. 1819. The Commission, other EU institutions and bodies, Member States’ authorities, beneficiary countries and/or regions establish systems. They use these systems to manage the risks to the budget, including the regularity of transactions. Examining systems is particularly useful for identifying recommendations for improvement. Each MFF heading, including revenue, involves many individual systems. We select a sample of systems each year. We present the results with recommendations for improvement. Grp_we_arrive_at_our_opinions_in_the_statement_of_assurance. 20. All our work reported in chapters 4 to 9 forms the basis for our opinion on the regularity of transactions underlying the European Union's consolidated accounts. Our opinion is set out in the statement of assurance. In forming our opinion we consider whether an error is pervasive. Our work allows us to arrive at an informed opinion as to whether errors in the population exceed or fall within the materiality limits. Our best estimate of the level of error for overall spending in 2015 is 3,8 %. We have more than 95 % confidence that the level of error for the audited population is material. The estimated level of error found in different MFF headings varies as described in chapters 5 to 9. We assessed error as pervasive — extending across the majority of spending areas. Grp_fraud. 21. If we have reason to suspect that fraudulent activity has taken place, we report this to OLAF, the Union’s antifraud office. OLAF is responsible for carrying out any resulting investigations. We report several cases per year to OLAF.

How we test transactions 4.5.6.7.8.9.10.

Within each specific assessment (chapters 5 to 9), we carry out direct tests of transactions on the basis of a representative sample of transactions. Our testing provides an estimate of the extent to which the transactions in the population concerned are irregular. Transaction testing involves examining each selected transaction. We determine whether or not the claim or payment was made for the purposes approved by the budget and specified in relevant legislation. We examine the calculation of the amount of the claim or payment (for larger claims based on a representative selection of the items on which subsidy is based). This involves tracing the transaction down from the budgetary accounts to the level of the final recipient (e.g. a farmer, the organiser of a training course, or a development aid project promoter). We test compliance at each level. There is an error when the transaction (at any level): For revenue, our examination of value added tax and gross national income-based own resources takes as a starting point the relevant macroeconomic aggregates on which these are calculated. We examine and take assurance from the Commission’s control systems for processing these until the contributions of the Member States have been received and recorded in the consolidated accounts. For traditional own resources, we examine the accounts of the customs authorities and the flow of duties until the amounts are received by the Commission and recorded in the accounts. For expenditure we examine payments when expenditure has been incurred, recorded and accepted. This examination covers all categories of payments (including those made for the purchase of assets). We do not examine advances at the point they are made. We examine advance payments when: Our audit sample is designed to provide an estimate of the level of error in the audited population as a whole. We examine larger claims or payments by selecting items (e.g. invoices of a project, parcels in a claim by a farmer, see paragraph 5) to audit within individual transactions using monetary unit sampling (MUS). Thus in case of examined items being part of a project or claim by a farmer, the error rate reported for these items does not constitute an error rate for the audited project or a claim by a farmer, but contributes to the overall evaluation of EU expenditure. We do not examine transactions in every Member State, beneficiary state and/or region each year. The examples provided in the annual report are for illustrative purposes and demonstrate the most typical errors found. The naming of certain Member States, beneficiary states and/or regions does not mean that the examples presented do not occur elsewhere. The illustrative examples presented in this report do not form a basis for conclusions to be drawn on the Member States, beneficiary states and/or regions concerned. Our approach is not designed to gather data on the frequency of error in the population. Therefore, figures presented on frequency of error are not an indication of the frequency of error in EU-funded transactions or in individual Member States. Our sampling approach applies different weighting to different transactions. Our sampling reflects the value of the expenditure concerned and the intensity of audit work. This weighting is removed in a frequency table which therefore gives as much weight to rural development as to direct support in the area of Natural Resources and to Social Fund expenditure as to regional and cohesion payments in the Cohesion chapter. The relative frequency of error in samples drawn in different Member States cannot be a guide to the relative level of error in different Member States. (1) is incorrectly calculated; or (2) does not meet a regulatory requirement or contractual provision. (3) the final recipient of EU funds (e.g. a farmer, a research institute, a company providing publicly procured works or services) provides evidence of their proper use; and (4) the Commission (or other institution or body managing EU funds) accepts that final use of funds is justified by clearing the advance payment.

is incorrectly calculated; or

does not meet a regulatory requirement or contractual provision.

the final recipient of EU funds (e.g. a farmer, a research institute, a company providing publicly procured works or services) provides evidence of their proper use; and

the Commission (or other institution or body managing EU funds) accepts that final use of funds is justified by clearing the advance payment.

How we evaluate and present the results of transaction testing 11.12.13.14.

Errors in transactions occur for a variety of reasons. They take a number of different forms depending on the nature of the breach and specific rule or contractual requirement not followed. Individual transactions may be wholly or partially affected by error. Errors detected and corrected before and independently of the checks carried out by us are excluded from the calculation and frequency of error. They demonstrate that the control systems have worked effectively. We consider whether individual errors are quantifiable or non-quantifiable. We take account of the extent to which it is possible to measure how much of the amount audited was affected by error. Many errors occur in the application of public procurement laws. To respect the basic principles of competition foreseen in EU laws and regulations, significant procedures must be advertised. Bids must be evaluated according to specified criteria. Contracts may not be artificially split to avoid breaching thresholds. Our criteria for quantification of public procurement errors are described in the document Non-compliance with the rules on public procurement — types of irregularities and basis for quantification. Our quantification may differ from that used by the Commission or Member States when deciding how to respond to the misapplication of the public procurement rules.

Estimated level of error 15.16.17.

We estimate the level of error using the most likely rate of error (MLE). We do this for each MFF heading, and for spending from the budget as whole. Only quantified errors are part of the calculation. The MLE percentage is a statistical estimate of the likely percentage of error in the population. Examples of errors are quantifiable breaches of applicable regulations, rules, and contract and grant conditions. We also estimate the lower error limit (LEL) and the upper error limit (UEL) (see illustration below). The percentage of the shaded area below the curve indicates the probability that the level of error of the population is between the LEL and the UEL. We plan our work on the basis of a materiality threshold of 2 %. We use the level of materiality as guidance for our opinion. We also take account of the nature, amount and context of errors when forming our opinion.

How we examine systems and report the results 18.19.

The Commission, other EU institutions and bodies, Member States’ authorities, beneficiary countries and/or regions establish systems. They use these systems to manage the risks to the budget, including the regularity of transactions. Examining systems is particularly useful for identifying recommendations for improvement. Each MFF heading, including revenue, involves many individual systems. We select a sample of systems each year. We present the results with recommendations for improvement.

How we arrive at our opinions in the statement of assurance 20.

All our work reported in chapters 4 to 9 forms the basis for our opinion on the regularity of transactions underlying the European Union's consolidated accounts. Our opinion is set out in the statement of assurance. In forming our opinion we consider whether an error is pervasive. Our work allows us to arrive at an informed opinion as to whether errors in the population exceed or fall within the materiality limits. Our best estimate of the level of error for overall spending in 2015 is 3,8 %. We have more than 95 % confidence that the level of error for the audited population is material. The estimated level of error found in different MFF headings varies as described in chapters 5 to 9. We assessed error as pervasive — extending across the majority of spending areas.

Suspected fraud 21.

If we have reason to suspect that fraudulent activity has taken place, we report this to OLAF, the Union’s antifraud office. OLAF is responsible for carrying out any resulting investigations. We report several cases per year to OLAF.

PART 3 Link between the audit opinions for the reliability of accounts and the regularity of transactions 22.23.

We have issued: Our work and our opinions follow the IFAC’s International Standards on Auditing and Codes of Ethics and INTOSAI’s International Standards of Supreme Audit Institutions. These standards also provide for the situation where auditors issue audit opinions on the reliability of accounts and the regularity of transactions underlying those accounts, by stating that a modified opinion on the regularity of transactions does not in itself lead to a modified opinion on the reliability of accounts. The financial statements, on which we express an opinion, recognise that there is a material issue in relation to breaches of the rules governing expenses charged to the EU budget. Accordingly, we have decided that the existence of a material level of error affecting regularity is not in itself a reason to modify our separate opinion on the reliability of the accounts. (a) an audit opinion on the consolidated accounts of the European Union for the financial year ended; and (b) audit opinions on the regularity of the revenue and payments underlying those accounts.

Link between the audit opinions for the reliability of accounts and the regularity of transactions 22.

(a) an audit opinion on the consolidated accounts of the European Union for the financial year ended; and

(b) audit opinions on the regularity of the revenue and payments underlying those accounts.

Anx_12_123456_differences_and_similarities_between_the_us_and_the_eca_models_for_estimating_the_level_of_error_in_budgetary. spending. A US act of Congress requires all government agencies to carry out sufficient testing of its own payments to calculate a statistically valid estimate of the annual amount of improper payments in those programmes wherever risk analysis suggests that such payments are likely to exceed 1,5 % of the total sum paid. Agencies must publish the result both as an extrapolated total and as a percentage of payments made. The auditors of government agencies are required to report on whether the process of producing and publishing these figures is in compliance with the relevant act of Congress. In the US (as in the EU for irregular payments) the definition of what is improper depends on the statute under which payment is made. The general definition for improper payments used in the US system is close to that for the irregular payments discussed in chapter 1 of this annual report. The table below outlines the similarities and differences between the US and EU approaches to testing and reporting on spending. The US improper payments total covers a broad and diverse range of public spending: from social welfare to defence. The areas of spending that dominate the EU budget (agriculture, infrastructure) form a relatively small part of the US Federal Budget. Some clear differences in patterns of error emerge from the breakdown of error types. Breaches of ancillary conditions (such as proper procurement and issues related to state aid that stem from the EU policy priority of ensuring a single market) have a greater impact on our estimated level of error than they do on the US improper payments total. Underpayments are a significant component of the US total, but not of our own. Documentation errors also have a larger impact on the US figures than they do in our estimated level of error. In preparing estimates of improper payments, US agencies are not allowed to rely on self-reporting by the recipients of the agency payments as the sole source basis for improper payments estimates, nor to net off the improper payments total by calculating the impact of recovery action. As described in paragraphs 1.42 to 1.43, corrective action taken before our audit reduces our estimate of error. The agencies concerned devote considerable resources to the examination of selected payments. For example, for the US Social Security Administration close to six thousand payments per annum are reviewed to calculate the improper payment rate, and that nearly ninety staff years are devoted to the task. The authorities with whom we discussed the issue considered that, as a result of the exercise, the agencies concerned obtain high quality information on patterns of improper payments, and are frequently able to design and implement suitable action to address risk of error. Improper payments reported for 2015 amounted to 136,7 billion USD (123,2 billion euro), equivalent to 4,4 % of spending by the Federal government. There are many differences in the nature of the spending covered, and in specific definitions of error. The key difference is that in the US bodies managing funds are required to produce their own, statistically-valid, estimate of the level of error. US Federal Government ECA Responsibility for calculating the estimated level of error Managers of funds must provide the statistical estimate of level of error External auditors calculate the statistical estimate of level of error Coverage All payments from the budget are subject to sampling Only spending on programmes considered risky by managers of funds are sampled Key errors Incorrect amounts paid to eligible recipients Payments made to ineligible recipients Payments made for ineligible goods or services Payments for goods or services not received Duplicate payments Payments for which insufficient or no documentation was found Basis for error calculation Testing of a sample of operations selected with the use of statistical methods Ex-post corrective actions (carried out by managers of funds) can reduce the estimated level of error Presentation of the estimated level of error Extrapolated error presented in cash terms Extrapolated error presented as percentage Results Average level 2012-2015 4,1 % 4,3 % Level for 2015 4,4 % 3,8 %

ANNEX 1.2 1.2.3.4.5.6. Differences and similarities between the US and the ECA models for estimating the level of error in budgetary spending US AND EU APPROACHES TO TESTING AND REPORTING ON IMPROPER AND IRREGULAR PAYMENTS

A US act of Congress requires all government agencies to carry out sufficient testing of its own payments to calculate a statistically valid estimate of the annual amount of improper payments in those programmes wherever risk analysis suggests that such payments are likely to exceed 1,5 % of the total sum paid. Agencies must publish the result both as an extrapolated total and as a percentage of payments made. The auditors of government agencies are required to report on whether the process of producing and publishing these figures is in compliance with the relevant act of Congress. In the US (as in the EU for irregular payments) the definition of what is improper depends on the statute under which payment is made. The general definition for improper payments used in the US system is close to that for the irregular payments discussed in chapter 1 of this annual report. The table below outlines the similarities and differences between the US and EU approaches to testing and reporting on spending. The US improper payments total covers a broad and diverse range of public spending: from social welfare to defence. The areas of spending that dominate the EU budget (agriculture, infrastructure) form a relatively small part of the US Federal Budget. Some clear differences in patterns of error emerge from the breakdown of error types. Breaches of ancillary conditions (such as proper procurement and issues related to state aid that stem from the EU policy priority of ensuring a single market) have a greater impact on our estimated level of error than they do on the US improper payments total. Underpayments are a significant component of the US total, but not of our own. Documentation errors also have a larger impact on the US figures than they do in our estimated level of error. In preparing estimates of improper payments, US agencies are not allowed to rely on self-reporting by the recipients of the agency payments as the sole source basis for improper payments estimates, nor to net off the improper payments total by calculating the impact of recovery action. As described in paragraphs 1.42 to 1.43, corrective action taken before our audit reduces our estimate of error. The agencies concerned devote considerable resources to the examination of selected payments. For example, for the US Social Security Administration close to six thousand payments per annum are reviewed to calculate the improper payment rate, and that nearly ninety staff years are devoted to the task. The authorities with whom we discussed the issue considered that, as a result of the exercise, the agencies concerned obtain high quality information on patterns of improper payments, and are frequently able to design and implement suitable action to address risk of error. Improper payments reported for 2015 amounted to 136,7 billion USD (123,2 billion euro), equivalent to 4,4 % of spending by the Federal government. There are many differences in the nature of the spending covered, and in specific definitions of error. The key difference is that in the US bodies managing funds are required to produce their own, statistically-valid, estimate of the level of error. US Federal Government ECA Responsibility for calculating the estimated level of error Managers of funds must provide the statistical estimate of level of error External auditors calculate the statistical estimate of level of error Coverage All payments from the budget are subject to sampling Only spending on programmes considered risky by managers of funds are sampled Key errors Incorrect amounts paid to eligible recipients Payments made to ineligible recipients Payments made for ineligible goods or services Payments for goods or services not received Duplicate payments Payments for which insufficient or no documentation was found Basis for error calculation Testing of a sample of operations selected with the use of statistical methods Ex-post corrective actions (carried out by managers of funds) can reduce the estimated level of error Presentation of the estimated level of error Extrapolated error presented in cash terms Extrapolated error presented as percentage Results Average level 2012-2015 4,1 % 4,3 % Level for 2015 4,4 % 3,8 %

Grp. 1. We take account of corrective measures and this impacts on the individual findings included in the table (see also paragraphs 1.42 to 1.43). We do not examine transactions in every Member State, beneficiary state, and/or region each year. For 2015 we did not examine any transactions in Luxembourg, Malta, Croatia and Estonia. Our approach is not designed to gather data on the frequency of error in the population. Therefore, figures presented on frequency of error are not an indication of the frequency of error in EU funded transactions or in individual Member States. The relative frequency of error in samples drawn in different Member States cannot be a guide to the relative level of error in different Member States. Member State (shared management) MFFH 1b — Cohesion MFFH 2 Natural Resources Shared management MFF headings 1b Cohesion and 2 Natural Resources (total) European Regional Development Fund (ERDF), Cohesion Fund (CF) and European Social Fund (ESF) Market and direct support Rural development, environment, climate action and fisheries Number of transactions Number of transactions affected by errors Number of transactions affected by: Number of transactions Number of transactions affected by errors Number of transactions affected by: Number of transactions Number of transactions affected by errors Number of transactions affected by: Number of transactions Number of transactions affected by errors Number of transactions affected by: Only OCI/NQE Quantifiable errors < 20 % 20 %-80 % 80 %-100 % Only OCI/NQE Quantifiable errors < 20 % 20 %-80 % 80 %-100 % Only OCI/NQE Quantifiable errors < 20 % 20 %-80 % 80 %-100 % Only OCI/NQE Quantifiable errors < 20 % 20 %-80 % 80 %-100 % BE Belgium 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 2 0 2 1 1 0 5 2 0 2 1 1 0 BG Bulgaria 9 0 0 0 0 0 0 4 2 0 2 2 0 0 5 0 0 0 0 0 0 18 2 0 2 2 0 0 CZ Czech Republic 28 6 2 4 2 0 2 4 0 0 0 0 0 0 5 4 0 4 4 0 0 37 10 2 8 6 0 2 DK Denmark 0 0 0 0 0 0 0 8 5 0 5 5 0 0 0 0 0 0 0 0 0 8 5 0 5 5 0 0 DE Germany 16 6 3 3 2 1 0 24 5 2 3 3 0 0 14 3 0 3 2 1 0 54 14 5 9 7 2 0 IE Ireland 0 0 0 0 0 0 0 8 0 0 0 0 0 0 0 0 0 0 0 0 0 8 0 0 0 0 0 0 EL Greece 17 2 2 0 0 0 0 8 1 0 1 1 0 0 5 2 0 2 1 0 1 30 5 2 3 2 0 1 ES Spain 17 7 5 2 0 2 0 20 5 0 5 4 0 1 21 8 3 5 3 2 0 58 20 8 12 7 4 1 FR France 8 3 0 3 2 0 1 32 10 1 9 7 2 0 14 11 8 3 2 1 0 54 24 9 15 11 3 1 IT Italy 25 9 5 4 1 2 1 20 8 1 7 5 2 0 22 13 7 6 4 1 1 67 30 13 17 10 5 2 CY Cyprus 0 0 0 0 0 0 0 4 3 0 3 3 0 0 0 0 0 0 0 0 0 4 3 0 3 3 0 0 LV Latvia 4 1 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 1 0 1 1 0 0 LT Lithuania 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 3 3 0 0 0 0 5 3 3 0 0 0 0 HU Hungary 17 6 2 4 1 3 0 8 3 0 3 3 0 0 5 1 0 1 1 0 0 30 10 2 8 5 3 0 NL Netherlands 0 0 0 0 0 0 0 4 0 0 0 0 0 0 0 0 0 0 0 0 0 4 0 0 0 0 0 0 AT Austria 8 2 1 1 0 1 0 0 0 0 0 0 0 0 10 0 0 0 0 0 0 18 2 1 1 0 1 0 PL Poland 41 16 11 5 1 2 2 8 0 0 0 0 0 0 22 4 3 1 1 0 0 71 20 14 6 2 2 2 PT Portugal 12 5 5 0 0 0 0 0 0 0 0 0 0 0 8 2 2 0 0 0 0 20 7 7 0 0 0 0 RO Romania 4 2 0 2 1 1 0 8 2 1 1 1 0 0 11 3 0 3 0 2 1 23 7 1 6 2 3 1 SI Slovenia 8 1 1 0 0 0 0 4 0 0 0 0 0 0 0 0 0 0 0 0 0 12 1 1 0 0 0 0 SK Slovak Republic 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 1 0 1 0 0 1 5 1 0 1 0 0 1 SF Finland 0 0 0 0 0 0 0 4 1 0 1 1 0 0 5 4 0 4 4 0 0 9 5 0 5 5 0 0 SE Sweden 0 0 0 0 0 0 0 4 2 0 2 2 0 0 0 0 0 0 0 0 0 4 2 0 2 2 0 0 UK United Kingdom 9 6 2 4 3 1 0 8 5 0 5 5 0 0 10 6 3 3 3 0 0 27 17 5 12 11 1 0 Total 223 72 39 33 14 13 6 180 52 5 47 42 4 1 172 67 29 38 26 8 4 575 191 73 118 82 25 11 Grp. 2. European Commission directorate-general (DG), other institutions and bodies (all type of management) Total number of transactions examined Number of transactions affected by one or more errors Number of transactions affected by: Only other compliance issues and non quantifiable errors Quantifiable errors Quantifiable errors < 20 % Quantifiable errors 20 %-80 % Quantifiable errors 80 %-100 % European Commission AGRI DG Agriculture and Rural Development 344 116 31 85 68 12 5 DEVCO DG International Cooperation and Development 49 14 5 9 7 2 0 HOME DG Migration and Home Affairs 8 4 2 2 2 0 0 JUST DG Justice and Consumers 1 0 0 0 0 0 0 SANTE DG Health and Food Safety 2 1 0 1 0 1 0 EACEA Education, Audiovisual and Culture Executive Agency 7 5 1 4 3 0 1 EMPL DG Employment, Social Affairs and Inclusion 82 30 16 14 8 5 1 JRC Joint Research Centre 3 2 2 0 0 0 0 NEAR DG Neighbourhood and Enlargement Negotiations 46 13 2 11 8 3 0 PMO Office for the Administration and Payment of Individual Entitlements 71 7 3 4 1 1 2 REGIO DG Regional and Urban Policy 150 43 23 20 7 8 5 ESTAT Eurostat 1 0 0 0 0 0 0 CONNECT DG Communications Networks, Content & Technology 20 9 2 7 4 2 1 MOVE DG Mobility and Transport 7 1 1 0 0 0 0 RTD DG Research and Innovation 55 35 15 20 14 5 1 ENER DG Energy 11 7 5 2 2 0 0 GROW DG Internal Market, Industry, Entrepreneurship and SMEs 7 1 1 0 0 0 0 EAC DG Education and Culture 32 13 7 6 5 1 0 ECFIN DG Economic and Financial Affairs 2 1 0 1 1 0 0 TAXUD DG Taxation and Customs Union 1 1 1 0 0 0 0 COMM DG Communication 1 0 0 0 0 0 0 FISMA DG Financial Stability, Financial Services and Capital Markets Union 1 0 0 0 0 0 0 ENV DG Environment 7 2 0 2 2 0 0 MARE DG Maritime Affairs and Fisheries 12 3 3 0 0 0 0 ECHO DG Humanitarian Aid and Civil Protection (ECHO) 36 7 1 6 6 0 0 OP Office for Publication 1 1 1 0 0 0 0 DIGIT DG Informatics 3 1 1 0 0 0 0 HR DG Human Resources and Security 5 0 0 0 0 0 0 OIB Office for Infrastructure and Logistics in Brussels 6 1 1 0 0 0 0 OIL Office for Infrastructure and Logistics in Luxembourg 2 0 0 0 0 0 0 SJ Office for Legal Service 1 1 0 1 1 0 0 SCIC Interpretation 1 0 0 0 0 0 0 Subtotal European Commission 975 319 124 195 139 40 16 Other instutions and bodies European Parliament 25 2 1 1 1 0 0 European External Action Service 25 9 7 2 2 0 0 Council of the European Union 9 0 0 0 0 0 0 Court of Justice of the European Union 6 1 1 0 0 0 0 European Court of Auditors 3 0 0 0 0 0 0 Other bodies 6 0 0 0 0 0 0 Subtotal other instutions and bodies 74 12 9 3 3 0 0 Total 1049 331 133 198 142 40 16

We take account of corrective measures and this impacts on the individual findings included in the table (see also paragraphs 1.42 to 1.43). We do not examine transactions in every Member State, beneficiary state, and/or region each year. For 2015 we did not examine any transactions in Luxembourg, Malta, Croatia and Estonia. Our approach is not designed to gather data on the frequency of error in the population. Therefore, figures presented on frequency of error are not an indication of the frequency of error in EU funded transactions or in individual Member States. The relative frequency of error in samples drawn in different Member States cannot be a guide to the relative level of error in different Member States. Member State (shared management) MFFH 1b — Cohesion MFFH 2 Natural Resources Shared management MFF headings 1b Cohesion and 2 Natural Resources (total) European Regional Development Fund (ERDF), Cohesion Fund (CF) and European Social Fund (ESF) Market and direct support Rural development, environment, climate action and fisheries Number of transactions Number of transactions affected by errors Number of transactions affected by: Number of transactions Number of transactions affected by errors Number of transactions affected by: Number of transactions Number of transactions affected by errors Number of transactions affected by: Number of transactions Number of transactions affected by errors Number of transactions affected by: Only OCI/NQE Quantifiable errors < 20 % 20 %-80 % 80 %-100 % Only OCI/NQE Quantifiable errors < 20 % 20 %-80 % 80 %-100 % Only OCI/NQE Quantifiable errors < 20 % 20 %-80 % 80 %-100 % Only OCI/NQE Quantifiable errors < 20 % 20 %-80 % 80 %-100 % BE Belgium 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 2 0 2 1 1 0 5 2 0 2 1 1 0 BG Bulgaria 9 0 0 0 0 0 0 4 2 0 2 2 0 0 5 0 0 0 0 0 0 18 2 0 2 2 0 0 CZ Czech Republic 28 6 2 4 2 0 2 4 0 0 0 0 0 0 5 4 0 4 4 0 0 37 10 2 8 6 0 2 DK Denmark 0 0 0 0 0 0 0 8 5 0 5 5 0 0 0 0 0 0 0 0 0 8 5 0 5 5 0 0 DE Germany 16 6 3 3 2 1 0 24 5 2 3 3 0 0 14 3 0 3 2 1 0 54 14 5 9 7 2 0 IE Ireland 0 0 0 0 0 0 0 8 0 0 0 0 0 0 0 0 0 0 0 0 0 8 0 0 0 0 0 0 EL Greece 17 2 2 0 0 0 0 8 1 0 1 1 0 0 5 2 0 2 1 0 1 30 5 2 3 2 0 1 ES Spain 17 7 5 2 0 2 0 20 5 0 5 4 0 1 21 8 3 5 3 2 0 58 20 8 12 7 4 1 FR France 8 3 0 3 2 0 1 32 10 1 9 7 2 0 14 11 8 3 2 1 0 54 24 9 15 11 3 1 IT Italy 25 9 5 4 1 2 1 20 8 1 7 5 2 0 22 13 7 6 4 1 1 67 30 13 17 10 5 2 CY Cyprus 0 0 0 0 0 0 0 4 3 0 3 3 0 0 0 0 0 0 0 0 0 4 3 0 3 3 0 0 LV Latvia 4 1 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 1 0 1 1 0 0 LT Lithuania 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 3 3 0 0 0 0 5 3 3 0 0 0 0 HU Hungary 17 6 2 4 1 3 0 8 3 0 3 3 0 0 5 1 0 1 1 0 0 30 10 2 8 5 3 0 NL Netherlands 0 0 0 0 0 0 0 4 0 0 0 0 0 0 0 0 0 0 0 0 0 4 0 0 0 0 0 0 AT Austria 8 2 1 1 0 1 0 0 0 0 0 0 0 0 10 0 0 0 0 0 0 18 2 1 1 0 1 0 PL Poland 41 16 11 5 1 2 2 8 0 0 0 0 0 0 22 4 3 1 1 0 0 71 20 14 6 2 2 2 PT Portugal 12 5 5 0 0 0 0 0 0 0 0 0 0 0 8 2 2 0 0 0 0 20 7 7 0 0 0 0 RO Romania 4 2 0 2 1 1 0 8 2 1 1 1 0 0 11 3 0 3 0 2 1 23 7 1 6 2 3 1 SI Slovenia 8 1 1 0 0 0 0 4 0 0 0 0 0 0 0 0 0 0 0 0 0 12 1 1 0 0 0 0 SK Slovak Republic 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 1 0 1 0 0 1 5 1 0 1 0 0 1 SF Finland 0 0 0 0 0 0 0 4 1 0 1 1 0 0 5 4 0 4 4 0 0 9 5 0 5 5 0 0 SE Sweden 0 0 0 0 0 0 0 4 2 0 2 2 0 0 0 0 0 0 0 0 0 4 2 0 2 2 0 0 UK United Kingdom 9 6 2 4 3 1 0 8 5 0 5 5 0 0 10 6 3 3 3 0 0 27 17 5 12 11 1 0 Total 223 72 39 33 14 13 6 180 52 5 47 42 4 1 172 67 29 38 26 8 4 575 191 73 118 82 25 11

European Commission directorate-general (DG), other institutions and bodies (all type of management) Total number of transactions examined Number of transactions affected by one or more errors Number of transactions affected by: Only other compliance issues and non quantifiable errors Quantifiable errors Quantifiable errors < 20 % Quantifiable errors 20 %-80 % Quantifiable errors 80 %-100 % European Commission AGRI DG Agriculture and Rural Development 344 116 31 85 68 12 5 DEVCO DG International Cooperation and Development 49 14 5 9 7 2 0 HOME DG Migration and Home Affairs 8 4 2 2 2 0 0 JUST DG Justice and Consumers 1 0 0 0 0 0 0 SANTE DG Health and Food Safety 2 1 0 1 0 1 0 EACEA Education, Audiovisual and Culture Executive Agency 7 5 1 4 3 0 1 EMPL DG Employment, Social Affairs and Inclusion 82 30 16 14 8 5 1 JRC Joint Research Centre 3 2 2 0 0 0 0 NEAR DG Neighbourhood and Enlargement Negotiations 46 13 2 11 8 3 0 PMO Office for the Administration and Payment of Individual Entitlements 71 7 3 4 1 1 2 REGIO DG Regional and Urban Policy 150 43 23 20 7 8 5 ESTAT Eurostat 1 0 0 0 0 0 0 CONNECT DG Communications Networks, Content & Technology 20 9 2 7 4 2 1 MOVE DG Mobility and Transport 7 1 1 0 0 0 0 RTD DG Research and Innovation 55 35 15 20 14 5 1 ENER DG Energy 11 7 5 2 2 0 0 GROW DG Internal Market, Industry, Entrepreneurship and SMEs 7 1 1 0 0 0 0 EAC DG Education and Culture 32 13 7 6 5 1 0 ECFIN DG Economic and Financial Affairs 2 1 0 1 1 0 0 TAXUD DG Taxation and Customs Union 1 1 1 0 0 0 0 COMM DG Communication 1 0 0 0 0 0 0 FISMA DG Financial Stability, Financial Services and Capital Markets Union 1 0 0 0 0 0 0 ENV DG Environment 7 2 0 2 2 0 0 MARE DG Maritime Affairs and Fisheries 12 3 3 0 0 0 0 ECHO DG Humanitarian Aid and Civil Protection (ECHO) 36 7 1 6 6 0 0 OP Office for Publication 1 1 1 0 0 0 0 DIGIT DG Informatics 3 1 1 0 0 0 0 HR DG Human Resources and Security 5 0 0 0 0 0 0 OIB Office for Infrastructure and Logistics in Brussels 6 1 1 0 0 0 0 OIL Office for Infrastructure and Logistics in Luxembourg 2 0 0 0 0 0 0 SJ Office for Legal Service 1 1 0 1 1 0 0 SCIC Interpretation 1 0 0 0 0 0 0 Subtotal European Commission 975 319 124 195 139 40 16 Other instutions and bodies European Parliament 25 2 1 1 1 0 0 European External Action Service 25 9 7 2 2 0 0 Council of the European Union 9 0 0 0 0 0 0 Court of Justice of the European Union 6 1 1 0 0 0 0 European Court of Auditors 3 0 0 0 0 0 0 Other bodies 6 0 0 0 0 0 0 Subtotal other instutions and bodies 74 12 9 3 3 0 0 Total 1049 331 133 198 142 40 16

CHAPTER 2 Budgetary and financial management

Txt. 1. THE COURT’S OBSERVATIONS INTRODUCTION This chapter gives an overview of key budgetary and financial management issues arising in 2015 and affecting the EU budget and balance sheet. These include overall levels of spending and of commitments made, the relationship between budgetary and financial accounts, financial instruments, and how the budget relates to the multi-annual financial framework (MFF). The MFF is a seven-year framework for the annual budget of the EU. It is defined in a Council Regulation and fixes an overall annual ceiling on payment and commitment appropriations. The current MFF sets limits on commitments of 1083 billion euro and payments of 1024 billion euro for the seven-year period 2014 to 2020. A new feature in this MFF is that unused amounts under the payment ceiling and under the commitments ceiling automatically increase the MFF ceilings of subsequent years. 2015 was the second year of the current MFF. More than three-quarters of operational spending went to schemes operating under the rules of the previous MFF (subsidies to farmers for 2014, reimbursement of claims for cohesion projects for the 2007-2013 operational programmes, research projects under the Seventh Framework Programme, which began in 2007). Within the new MFF, the Commission disbursed 9,6 billion euro in pre-financing payments on new operational programmes in cohesion. The EU general budget does not itself deliver most of the benefits of EU membership (such as access to the single market), nor record all of the costs (such as co-financing and administering shared management spending). As we have noted, several financial mechanisms supporting EU policies are not financed directly from the EU budget nor recorded in the EU balance sheet. These include the European Financial Stability Facility, the European Stability Mechanism, the Single Resolution Mechanism and the European Investment Bank (EIB) and the (linked) European Investment Fund (EIF). For some of these we are not the auditors, limiting the extent to which they are covered in this chapter. Other mechanisms are partially recorded in the EU balance sheet such as the blending facilities and the European Fund for Strategic Investments (EFSI) (see paragraphs 2.32 to 2.37). One mechanism that we audit separately is the European Development Fund, for which our observations on budgetary and financial management are included in our specific annual report. OBSERVATIONS Outstanding commitments rose and future payment requirements continued to grow The Commission made more commitments in 2015 than in any previous year The Commission recorded budgetary commitments for a total 177,2 billion euro, representing 97,7 % of the total level available. A commitment has a different basis in different areas of the budget (see Figure 2.1). Commitments Budgetary commitments are required for a variety of contracts, decisions or agreements made between the EU and individuals, companies, organisations, institutions, agencies or governments. Examples are: Source: European Court of Auditors. Headings of the MFF Typical basis for commitment Total commitments made in 2015 (billion euro) Heading 1.a — Competitiveness Contracts with researchers and research organisations, organisations handling space programmes, national agencies handling ERASMUS programmes. 18,9 Heading 1.b — Cohesion Annual instalments for planned cohesion spending in each Member State over the MFF. 69,5 Heading 2 — Natural resources Agreements with farmers and agri-business imposing conditions over more than one year in the area of rural development. Fisheries agreements with third countries to secure fishing rights. 67,4 Heading 3 — Security and citizenship Contracts and agreements to promote health, culture, asylum, equality and justice. 2,8 Heading 4 — Global Europe Financing agreements signed with development partners for a specific period. 9,4 Heading 5 — Administration Procurement contracts for supplies and services. 9,2 THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Commitment appropriations available in 2015 were higher than in any previous year. Two decisions, permitted under EU budgetary rules and linked to the delay in approval of programmes of the current MFF, contributed to this: (i) a Council decision to transfer 16,5 billion euro of unused commitments from 2014 to 2015, increasing the MFF ceiling to 163 billion euro; (ii) a Commission decision to carry over 12,1 billion euro of 2014 unused commitment appropriations, of which 11 billion euro were near completion. Txt. 2. This increase in commitment appropriations for 2015 was expected, as it included those additional commitments linked to Cohesion policy which were supposed to be made in 2014, but have been postponed in line with the applicable legal framework (due to delays in adopting operational programmes under Cohesion). Further increases in commitment appropriations available resulted from assigned revenue (6,8 billion euro) and other amending budgets (0,5 billion euro). Total commitment appropriations available therefore amounted to 181,3 billion euro, equivalent to 1,3 % of EU GNI. Only 2 billion euro of payment appropriations were available for carryover to 2015. The transfer of commitment appropriations therefore increases the pressure on future payment budgets. With most of these commitments being subject to an N + 3 rule, it is normal that these additional commitments do not lead to a related increase in payments. Instead they lead indeed to an increase in the level of outstanding payments RAL (reste à liquider). Outstanding commitments are those commitments that have been made but not cleared by payment or cancelled. These reached 217,7 billion euro (see Figure 2.2). Liabilities and accrued costs are reflected in the balance sheet without necessarily requiring a commitment (for example, for pensions earned by EU civil servants, or the likely cost of decommissioning EU nuclear facilities). Liabilities and accrued costs not covered by a budgetary commitment add a further 121 billion euro, giving a total to be funded from future budgets of 339 billion euro at 31 December 2015 (2014: 305 billion euro). As we have noted in previous years, the Commission does not prepare and publish an annually updated cash flow forecast, spanning a seven to ten year time horizon, covering budgetary ceilings, payment needs, capacity constraints and potential cancellation of commitments. See the Commission’s reply to paragraph 2.40. Evolution of accumulated outstanding commitments (billion euro) Source: Consolidated annual accounts of the European Union — Financial years 2007-2015. THE COURT’S OBSERVATIONS The initial budget for payments in 2015 was higher than in previous years … The EU plans and reports the budget for payments on a cash basis and must balance receipts and payments: the budget cannot be funded by borrowing. The initial EU budget for payments in 2015 was higher than in any previous year. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES … and payments on assigned revenue and appropriations carried over mean that final payments were higher than the MFF ceiling The MFF Regulation adopted in 2013 set a maximum level for payment appropriations for 2015 at 142 billion euro. The initial budget was set just below this ceiling at 141,2 billion euro. The final level of payments was 4 billion euro higher than the level set in the original budget (see paragraph 2.2 and Figure 2.4), and was second only to 2013. For the third successive year the final level of payments was higher than the MFF ceiling for the year. Figure 2.3 shows the gap between the MFF ceiling and final payments over the last seven years. The Court is comparing the voted budget with the level of payments including: (3) payments made on appropriations generated by assigned revenue (these are on top of the payment appropriations authorised in the 2015 budget and the MFF ceiling) and (4) payments related to appropriations carried over from previous years (which are to be counted against the MFF payment ceiling of the year, when the appropriations had been authorised by the European Parliament and the Council in the first place). Txt. 3. Final payments compared to the MFF ceiling (billion euro) Source: Consolidated annual accounts of the European Union — Financial years 2009-2015 and MFF Regulations. Payment appropriations and payments made in 2015 (billion euro) Source: Consolidated annual accounts of the European Union — Financial year 2015, Aggregated reports on the implementation of the budget and explanatory notes, tables 5.1 and 5.3. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Parliament and Council approved eight amending budgets. The first was the biggest and dealt with the reprogramming of 16,5 billion euro (paragraph 2.6). The other amending budgets together increased available commitment appropriations by 486 million euro and that available payment appropriations by 85 million euro. Notably they: (5) re-allocated 1,4 billion euro of commitment appropriations to the EFSI Guarantee Fund, largely from the Connecting Europe Facility (see paragraph 2.35); (6) added 407 million euro of commitment appropriations to the budget for migration; and (7) re-allocated 127 million euro of payment appropriations to the migration budget from other headings. Txt. 4. Cost claims for cohesion fell Authorities in the Member States presented a lower level of cost claims for reimbursement for cohesion activities in 2015 than in 2014 (42,2 billion euro in 2015; 2014, 56,6 billion euro). This led to a fall in the level of unpaid cost claims from 23,2 billion euro at the end of 2014 to 10,8 billion euro at the end of 2015. Claims received in the last month of the year nearly halved to 10,5 billion euro (2014, 19,4 billion euro), as did claims received in the last four days of the year (2015, 5,4 billion euro; 2014, 10,1 billion euro). Of the unpaid total of 10,8 billion euro at year end, 2,8 billion euro had remained unpaid since the end of 2014. The EU budget has provided pre-financing to all Member States for both the 2007-2013 and 2014-2020 MFFs. In total uncleared pre-financing amounts to 32,7 billion euro at the end of 2015. Thus the unpaid bills total is not an indication that Member States are overall owed funds from the budget. The Common Provisions Regulation 2014-2020 sets out that the programming period’s pre-financing shall be used for financing of operations carried out under the programming period in question. The period from initial commitment to acceptance of expenditure continues to be long Figure 2.5 provides a breakdown by age of outstanding commitments and pre-financing. The average length of time from commitment to payment is 2,4 years (2,1 years in 2014). In most cases, the first payment on a programme or project (reducing the level of outstanding commitment) consists of pre-financing. Pre-financing remains outstanding until the recipient of EU funds delivers the service or justifies the use of EU funding. The outstanding amount (stock) of commitments is equivalent to nearly two and half times the amount recorded in one year (the flow). Gross outstanding pre-financing is equivalent to just over half a year of payments. The combined average period from commitment to acceptance of expenditure amounts to three years across the EU budget. Use of pre-financing is much more pronounced for indirect management/direct management areas of spending (where the stock of pre-financing is equivalent to 1,6 years payments) than for shared management. For heading 4 the total amount of commitments and of pre-financing outstanding is equivalent to 4,6 years of normal spending. The Commission considers that the reste à liquider (RAL or outstanding commitments) is consistent with the relevant rules for automatic de-commitments. Lengthy delays between initial commitment and final clearing of expenditure is linked to problems of missing documentation at closure and reduces flexibility. As we note in paragraph 2.3 payments in 2015 frequently continued to relate to programmes of previous periods. For cohesion the final date of eligibility is 31 December 2015. Closure documents must be submitted by 31 March 2017. Regulation (EU) No 1083/2006 foresees that a share of 5 % of the total EU allocation is retained until closure of the programme. The co-legislators decided to amend the regulation and extend the N+2 automatic de-commitment deadline for two Member States. This has reduced the pressure on these Member States to submit payment claims within a shorter reference period. Additionally, Croatia’s accession treaty allows an exception to the automatic de-commitment rule. It also provides for N + 3. Finally, a taskforce for better implementation was set up in 2014 to help Member States optimise the use of the 2007-2013 Cohesion policy envelopes. Progress is being monitored by the Commission. Breakdown of the Commission’s outstanding budgetary commitments and pre-financing (million euro) Outstanding budgetary commitments Heading < 2009 2009 2010 2011 2012 2013 2014 2015 Total Years of commitments 1.a Competitiveness 295 672 1209 1541 3885 5942 7844 13066 34455 2,1 1.b Cohesion 1653 38 237 1455 7191 29459 21324 65016 126372 2,6 2 Natural resources 223 62 82 127 213 7231 2140 18112 28191 1,8 3 Security and citizenship 21 39 62 136 277 580 350 1671 3137 1,5 4 Global Europe 938 522 883 1412 3364 4719 5390 7446 24673 3,0 5 Administration 0 0 0 0 0 0 0 295 295 0,0 Total 3130 1333 2473 4671 14931 47931 37049 105606 217123 2,4 Pre-financing Heading < 2009 2009 2010 2011 2012 2013 2014 2015 Total Years of payments 1.a Competitiveness 127 165 422 1382 3353 6117 5343 8636 25545 1,5 1.b Cohesion 12934 9347 689 69 33 75 2042 7523 32711 0,6 2 Natural resources 3443 131 60 78 120 172 393 2397 6795 0,1 3 Security and citizenship 19 27 58 92 192 675 792 1430 3284 1,6 4 Global Europe 236 461 397 817 947 1634 2520 5446 12458 1,6 5 Administration 0 0 0 0 0 0 0 2 2 0,0 Total (gross pre-financing) 16758 10130 1626 2439 4645 8673 11090 25434 80795 0,6 THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Backlogs in the use of European Structural and Investment (ESI) funds are significant By the end of 2015 payments to Member States for the 2007-2013 MFF ESI funds had reached 400,8 billion euro (90 % of the total of 446,2 billion euro for all approved operational programmes). We provide an analysis of the extent to which the Commission has made payments against commitments for each Member State in Figure 2.6. Five Member States account for more than half of the unused commitments of ESI funds that have not led to payment. The five Member States indicated by the Court account for the most significant ESI allocations. In relation to the issue of low absorption levels, the taskforce for better implementation set up by the Commission is actively involved in assisting Member States most concerned by low absorption levels. In some Member States the unclaimed EU contribution, together with required national co-financing, is equivalent to a significant share of total general government expenditure (see Figure 2.7). In seven Member States the total accumulated share that could be claimed from EU funds (taking both the 2007-2013 and 2014-2020 periods) is more than 15 % of annual general government expenditure. A task force for better implementation in Member States was set up in 2014 to improve implementation and budget absorption in the Member States with the lowest absorption and proportionally highest outstanding EU contributions. In order to use all the funds available under the 2007-2013 MFF, authorities in the Member States will need to submit valid claims equalling the sum of all outstanding commitments together with required national co-financing and the amount of pre-financing already provided from the budget. Figure 2.8 shows the relative shares of unused 2007-2013 commitments for each Member State together with the pre-financing paid for the 2014-2020 MFF at the end of 2015. Absorption in % and totals of 2007-2013 MFF ESI funds at 31 December 2015 Croatia acceded to the EU in July 2013. Absorption figures for Croatia are therefore not comparable to those of other Member States. Source: European Court of Auditors calculation based on Commission accounting data. Unused commitments of ESI funds at 31 December 2015 as a percentage of 2015 general government expenditure Source: European Court of Auditors based on information from the Commission. Eurostat data on general government expenditure for 2015, April 2016. 2007-2013 ESI Funds unused commitments and 2014-2020 pre-financing at the end of 2015 (million euro) Source: European Court of Auditors based on information from the Commission. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Financial instruments form a growing part of the budget … Financial instruments are principally composed of loans, equity instruments, guarantees and risk sharing instruments. Financial instruments under indirect management In the 2014-2020 MFF, financial instruments under indirect management have gained in importance and form a growing part of disbursements from the EU budget. The EIB Group manages almost all of the financial instruments under indirect management. This includes the Guarantee Fund for External Actions (see Figure 2.9), which represents 32 % of the total assets held in financial instruments. Only a few very small instruments (assets of 74 million euro), and some technical assistance facilities set up prior to 2014, are managed by other international financial institutions. For the 2014-2020 MFF the Commission has signed only two delegation agreements with bodies outside the EIB Group. The Guarantee Fund for External Action is not a financial instrument in the sense of the Financial Regulation. It pre-existed the Financial Regulation and is based on a lex specialis, Council Regulation (EC, Euratom) No 480/2009 of 25 May 2009 establishing a Guarantee Fund for external actions. Total assets of 6,5 billion euro (1) at 31 December 2015 in financial instruments under indirect management by type of instrument See the table on financial instruments in the FSDA of the accounts. Source: Consolidated annual accounts of the European Union — Financial year 2015 (FSDA section on financial instruments) and compilation of data from financial instruments’ 2015 reporting packages. THE COURT’S OBSERVATIONS The Commission is closing financial instruments from the previous MFFs or merging them into new ones. In the 2014-2020 MFF: (8) expected payments into financial instruments under indirect management almost double to 7,4 billion euro (2007-2013 MFF: 3,8 billion euro); (9) the EU budget provides an additional guarantee of up to 16 billion euro to the newly created EFSI (see paragraph 2.33); (10) the EU Initiative for Small and Medium-sized Enterprises (SME Initiative) (originally designed as a crisis-response instrument) combines EU budget resources available under COSME and Horizon 2020 (175 million euro each), and ERDF-EAFRD (8,5 billion euro) with EIB Group resources. Txt. 5. At the end of 2015, cumulative payments into the new instruments amounted to 1,3 billion euro (2014: 0,7 billion euro) as shown in the Figure 2.10. Of this: (11) guarantee instruments amount to 1,1 billion euro (88 %) of cumulative payments; (12) equity instruments amount to 0,2 billion euro (12 %) of cumulative payments. Txt. 6. 2014-2020 MFF financial instruments under indirect management Source: SWD(2015) 206 final, Delegation Agreements and ABAC Workflow. (million euro) Name of the instrument Type of instrument Indicative budget 2014-2020 Cumulative commitments Cumulative payments 2014 2015 2014 2015 HORIZON 2020 — InnovFin Loan and Guarantee Service for RDI — EIB Guarantee 1060 483 646 483 646 of which earmarked revenue re-allocated from RSFF 375 430 374 395 HORIZON 2020 — InnovFin SMEs & Small Midcaps R&I Loans Service — EIF Guarantee 1060 169 306 169 304 HORIZON 2020 — InnovFin Equity Facility for R&I — EIF Equity 430 0 125 0 110 CEF Debt Instrument — EIB Guarantee 2557 128 247 0 0 CEF Equity Instrument Equity 0 10 0 0 COSME Equity Facility for Growth — EIF Equity 633 52 102 33 41 COSME Loan Guarantee Facility — EIF Guarantee 686 89 238 42 131 Student Loan Guarantee Facility (Erasmus +) — EIF Guarantee 517 28 61 0 20 Employment and Social Innovation (EaSI) — EIF Guarantee 193 28 60 0 10 Cultural and Creative Sector Guarantee Facility Guarantee 123 0 0 0 0 Private Finance for Energy Efficiency Instrument (PF4EE) — EIB Guarantee 80 30 50 6 12 Natural Capital Financing Facility (NCFF) — EIB Guarantee 60 10 30 3 12 Total 7399 1017 1875 735 1286 THE COURT’S OBSERVATIONS … and of the balance sheet … Total assets held in financial instruments under indirect management increased by 650 million euro (11 %) on the previous year (see Figure 2.11). Key elements of this increase are: (13) a 251 million euro increase in investments in debt securities and fixed income securities resulting from transfers from the EU budget to fiduciary accounts; (14) a 177 million euro increase in investments in euro-denominated bonds by the Guarantee Fund for External Actions, to which 144 million euro had been transfered from the EU budget. Txt. 7. Total assets of financial instruments under indirect management by type of asset (billion euro) Source: Consolidated annual accounts of the European Union — Financial years 2014 and 2015. THE COURT’S OBSERVATIONS Cash represents 20 % of financial instruments’ total assets, and remained stable at 1,3 billion euro at the end of 2015 (2014: 1,3 billion euro) (see Figure 2.12). Cash and cash equivalents belonging to financial instruments (1) See Note 2.8 of the 2015 accounts. Source: Compilation of data from financial instruments’ 2015 reporting packages. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES During 2015, 211 million euro were returned from the financial instruments’ trust accounts managed by the entrusted entities: 126 million euro to the general budget and 85 million euro as internal assigned revenue. Of this: (15) 134 million euro represent revenues and repayments; (16) 59 million euro represent returns to the general EU budget of unused funds held on the trust accounts of financial instruments and technical assistance facilities at closure and 18 million euro represent cancelled European Neighbourhood and Partnership Instruments (ENPI) operations. Txt. 8. For the financial year 2015, total administrative and management fees for financial instruments amount to 63 million euro (2014: 50 million euro). For the new instruments cumulative fees for 2014 and 2015 amount to 72 million euro (5,77 % of total assets held). … and unutilised amounts in financial instruments under shared management remain relatively high According to the latest available report on financial instruments under shared management, describing the situation as at the end of 2014, 6,8 billion euro of unutilised amounts remained in these instruments. Of this 80 % was concentrated in five Member States. Italy (3 billion euro) constituted 45 % of the total. During our examination of the balance sheet we asked the Commission to provide additional analysis on the use of these instruments in Italy up to 31 December 2015. This shows that 1,7 billion euro remained unutilised at 31 December 2015 for Italy. The Commission requested additional information from Italy by 31 December 2015. This information shows that 45 % for Italy remained unutilised as at 31 December 2015. The data obtained, checked against the bookings made in the 2015 Commission accounts, have been transmitted to the Court. Following the analysis, it appeared that there is no need to change the current methodology. However, implementation will be reassessed for the 2016 exercise. The first operations were recorded for the European Fund for Strategic Investments (EFSI) … The Commission (representing the EU) and the EIB signed the EFSI Agreement on 22 July 2015. EFSI is part of the Investment Plan for Europe announced in 2014 by the President of the European Commission. It aims to support investments in areas such as infrastructure, education, research and innovation, as well as risk finance for small businesses. The EU provides a budget guarantee of up to 16 billion euro (of which 8 billion euro will be transferred gradually to a specific fund between 2016 and 2022) and the EIB contributes 5 billion euro from its own resources to a risk reserve. The Commission and the EIB expect this to support EIB/EIF financing of up to 61 billion euro, with the objective of leveraging 315 billion euro of investments. The EU guarantees Infrastructure and Innovation Window (IIW) operations, implemented by the EIB, and Small and medium-sized enterprises Window (SMEW) operations, implemented by the EIF. Each of these has a debt portfolio (EU guarantee of 11 billion euro for the IIW and 1250 million euro for the SMEW) and an equity portfolio (EU guarantee of 2500 million euro for the IIW and 1250 million euro for the SMEW). The Commission will transfer funds from the guarantee to the EIB on receipt of the EIB’s request. To meet such claims the Commission set up a guarantee fund (EFSI guarantee fund). The EU budget will finance this fund over the period to 2020 by reducing planned spending on the Connecting Europe Facility (CEF) by 2,8 billion euro and on Horizon 2020 by 2,2 billion euro and by using 3 billion euro from unallocated margins under the MFF expenditure ceiling. In December 2015 the Commission committed 1,4 billion euro for payments into the fund. It made no payments in 2015. The funds will be paid to the EIB if required to honour guarantee calls in line with the EFSI Regulation under the conditions specified in the EFSI agreement. The EIB signed eight IIW debt portfolio operations (totalling 1092,7 million euro) of which it disbursed 193,8 million euro in 2015. The EIB signed two IIW equity portfolio operations (totalling 125 million euro in 2015), half of which is guaranteed by the EU. Of the amounts covered by the EU guarantee, the EIB disbursed only 7,6 million euro in 2015. The total EU guarantee on disbursed amounts and accrued interest on this portfolio amounted to 202 million euro at year end. Disbursements increase gradually when building up a portfolio of operations. The EFSI portfolio is currently being expanded with a considerable increase in size foreseen by the end of 2016. For the SME window, during 2015 the EIF used the EU guarantee only to guarantee operations under existing SME financial instruments. … while launch of the Connecting Europe Facility (CEF) was delayed The Commission and the EIB signed the CEF debt instrument agreement on 22 July 2015, the same day on which they signed the EFSI agreement. Merger of the LGTT and PBI portfolios was postponed until 2016. The Commission paid no contribution to the instrument in 2015 and the EIB signed no operations. CONCLUSIONS AND RECOMMENDATIONS The conclusions for 2015 The proliferation of financial mechanisms which are not directly funded by the EU budget nor audited by the Court poses risks both for accountability and the coordination of EU policies and operations (paragraph 2.4). The Commission made a high level of commitments in 2015, in part on appropriations brought forward from 2014. Outstanding commitments returned to previous levels (paragraphs 2.5 to 2.9). Amounts to be funded from future budgets rose to 339 billion euro. The Commission has not produced a cash flow forecast covering the next seven to ten years (paragraph 2.10). Such a forecast would enable stakeholders to anticipate future payment requirements and budgetary priorities and constructively plan ahead. The MFF is the tool ensuring medium- to long-term stability and predictability of future payment requirements and budgetary priorities. Therefore long-term forecasts can only be established taking into consideration existing financial frameworks. The payment implications of future financial frameworks which have not yet been agreed cannot be predicted with precision. The forecast for the post-2020 period will be presented in 2017 (as foreseen in point 9 of the Inter-institutional Agreement on budgetary discipline, on cooperation in budgetary matters and on sound financial management) and it will be an integral part of the Commission proposals for the post-2020 MFF (to be presented before 1 January 2018 based on Article 25 of the MFF Regulation). The Commission already committed itself to presenting a medium-term payment forecast (in the framework of the Mid-term review of the financial framework) which will analyse the sustainability of the payment ceilings and evolution of RAL and de-commitments until 2020. Cost claims submitted for cohesion during 2015 fell. This led to a reduction in cost claims awaiting reimbursement, but did not reduce the level of payments made (see paragraphs 2.14 and 2.15). Continuing lengthy delays between initial commitments and final clearings of expenditure means that patterns of spending are slow to respond to changing budgetary priorities and increases the risk that documentation is unavailable at closure (see paragraphs 2.16 and 2.17. The delay between commitments and payments are just reflecting the duration of ongoing projects. Some projects are large infrastructure projects in the Union or projects in rather difficult third countries and an average delay of 2,4 years should not be seen as particularly abnormal. The legal framework obliges the beneficiaries to keep the relevant documentation for a certain period after the closure. Outstanding commitments (together with outstanding pre-financing) in ESI Funds of the 2007-2013 MFF for some Member States remain high and frequently exceed 15 % of general government expenditure (paragraphs 2.19 to 2.21). Increasing amounts are held in financial instruments under indirect management, for such the EIB Group remains the preferred financial institution for the management (paragraph 2.23). Unused amounts of the financial instruments under shared management remain high and are concentrated in some Member States (paragraph 2.31). See reply to paragraph 2.31. Establishment of EFSI affected the launch of the CEF programme. Although it affects the assumptions the Commission made in its ex-ante assessment of the use of financial instruments under CEF, this was not re-evaluated. EFSI will also impact upon the use made of some other financial instruments (see paragraphs 2.32 to 2.38). The adoption of EFSI did not affect the overall objectives of CEF, but created an additional mechanism for the EU financial assistance, i.a. to infrastructural projects, in particular via debt finance. To take into account the launch of EFSI, the CEF DI Steering Committee in its first meeting defined the main principles for providing CEF and EFSI support, with a view to enhance their complementarity. Moreover, an addendum to the CEF ex ante assessment is under preparation to reflect (i) the latest analysis on the potential to launch a CEF Equity Instrument in the CEF sectors, and (ii) the market developments post-EFSI launch. This addendum is to be understood as complementing the previous ex ante assessment, and not as a revision. The coherence of CEF debt instrument with other EU initiatives, including EFSI, will be assessed in the mid-term review of the CEF. Recommendations We recommend that: (17) Recommendation 1: the Commission should act to reduce outstanding commitments by, for example, quicker decommitment, where appropriate, faster closure of the 2007-2013 programmes, where appropriate, wider use of net correction in cohesion, reducing cash held by fiduciaries and the compilation of payment plans and forecasts in those areas where outstanding commitments and other obligations are significant (see paragraphs 2.5 to 2.9, 2.19 and 2.28); Txt. 9. The Commission accepts the recommendation. Any measures taken by the Commission vis-à-vis the Member States, including efforts to reduce outstanding commitments and make wider use of net corrections in cohesion must obviously comply with the regulatory framework (18) Recommendation 2: the Commission should prepare and publish an annually updated cash flow forecast, spanning a seven to ten year time horizon, covering budgetary ceilings, payment needs, capacity constraints and potential decommitments (paragraph 2.10); Txt. 10. The Commission accepts the recommendation. The forecast for the post-2020 period will be presented in 2017 (as foreseen in point 9 of the Inter-institutional Agreement on budgetary discipline, on cooperation in budgetary matters and on sound financial management) and it will be an integral part of the Commission proposals for the post-2020 MFF (to be presented before 1 January 2018 based on Article 25 of the MFF Regulation). In addition the Commission will present a mid-term payment forecast assessing the sustainability of the current payment ceilings until the end of the current MFF in the framework of the Mid-term review of the MFF. (19) Recommendation 3: the Commission should consider in its budgetary and financial management the capacity constraints in certain Member States in order to ensure funds are used efficiently (see paragraphs 2.19 to 2.21); Txt. 11. The Commission accepts the recommendation within the framework set out by the MFF and the relevant legislation. The Commission has already put in place a Task Force for Better Implementation in order to improve the implementation of ESI funds in certain Member States, while ensuring full respect of legality and regularity of expenditure (see the Commission’s reply to observations 2.19 and 2.20). The Commission intends to ensure that full benefit is drawn from the work of the Task Force and will focus on delivering a proactive and targeted approach to Member States in order to support the implementation of the Cohesion policy funding in 2014-2020. (20) Recommendation 4: the Commission should recover unused cash balances in financial instruments under shared management (see paragraph 2.31) and remaining unused funds in indirect management financial instruments from previous MFFs for which the eligibility period has expired; Txt. 12. The Commission accepts the recommendation. For the indirectly managed financial instruments, a procedure for return and de-commitments of funds to the EU covering the transfer of GIF revenues and repayments to H2020 and COSME was finalised by the Commission with the EIF in November 2015 and execution is under way. It will be accompanied by monitoring visits to the EIF. For the shared management, any unused financial instruments allocations will be deducted at closure in accordance with the legal provisions for shared management programmes. (21) Recommendation 5: the Commission should re-evaluate the ex-ante assessment for the CEF debt instrument in the light of creation of the EFSI, and consider the impact of EFSI on other EU programmes and financial instruments. Txt. 13. The Commission partially accepts the recommendation. The coherence of CEF debt instrument with other EU initiatives, including EFSI, will be assessed in the mid-term review of the CEF. An addendum to the CEF ex ante assessment is under preparation to reflect (i) the latest analysis on the potential to launch a CEF Equity Instrument in the CEF sectors, and (ii) the market developments post-EFSI launch. This addendum is to be understood as complementing the previous ex ante assessment, and not as a revision The additionality of CEF financial instruments towards EFSI is based on a principle that CEF financing focuses on innovative, demonstrator and pilot projects and initiatives eligible under the relevant regulations in the CEF sectors, while the scope and eligibility criteria of EFSI support are broader.

THE COURT’S OBSERVATIONS INTRODUCTION This chapter gives an overview of key budgetary and financial management issues arising in 2015 and affecting the EU budget and balance sheet. These include overall levels of spending and of commitments made, the relationship between budgetary and financial accounts, financial instruments, and how the budget relates to the multi-annual financial framework (MFF). The MFF is a seven-year framework for the annual budget of the EU. It is defined in a Council Regulation and fixes an overall annual ceiling on payment and commitment appropriations. The current MFF sets limits on commitments of 1083 billion euro and payments of 1024 billion euro for the seven-year period 2014 to 2020. A new feature in this MFF is that unused amounts under the payment ceiling and under the commitments ceiling automatically increase the MFF ceilings of subsequent years. 2015 was the second year of the current MFF. More than three-quarters of operational spending went to schemes operating under the rules of the previous MFF (subsidies to farmers for 2014, reimbursement of claims for cohesion projects for the 2007-2013 operational programmes, research projects under the Seventh Framework Programme, which began in 2007). Within the new MFF, the Commission disbursed 9,6 billion euro in pre-financing payments on new operational programmes in cohesion. The EU general budget does not itself deliver most of the benefits of EU membership (such as access to the single market), nor record all of the costs (such as co-financing and administering shared management spending). As we have noted, several financial mechanisms supporting EU policies are not financed directly from the EU budget nor recorded in the EU balance sheet. These include the European Financial Stability Facility, the European Stability Mechanism, the Single Resolution Mechanism and the European Investment Bank (EIB) and the (linked) European Investment Fund (EIF). For some of these we are not the auditors, limiting the extent to which they are covered in this chapter. Other mechanisms are partially recorded in the EU balance sheet such as the blending facilities and the European Fund for Strategic Investments (EFSI) (see paragraphs 2.32 to 2.37). One mechanism that we audit separately is the European Development Fund, for which our observations on budgetary and financial management are included in our specific annual report. OBSERVATIONS Outstanding commitments rose and future payment requirements continued to grow The Commission made more commitments in 2015 than in any previous year The Commission recorded budgetary commitments for a total 177,2 billion euro, representing 97,7 % of the total level available. A commitment has a different basis in different areas of the budget (see Figure 2.1). Commitments Budgetary commitments are required for a variety of contracts, decisions or agreements made between the EU and individuals, companies, organisations, institutions, agencies or governments. Examples are: Source: European Court of Auditors. Headings of the MFF Typical basis for commitment Total commitments made in 2015 (billion euro) Heading 1.a — Competitiveness Contracts with researchers and research organisations, organisations handling space programmes, national agencies handling ERASMUS programmes. 18,9 Heading 1.b — Cohesion Annual instalments for planned cohesion spending in each Member State over the MFF. 69,5 Heading 2 — Natural resources Agreements with farmers and agri-business imposing conditions over more than one year in the area of rural development. Fisheries agreements with third countries to secure fishing rights. 67,4 Heading 3 — Security and citizenship Contracts and agreements to promote health, culture, asylum, equality and justice. 2,8 Heading 4 — Global Europe Financing agreements signed with development partners for a specific period. 9,4 Heading 5 — Administration Procurement contracts for supplies and services. 9,2 THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Commitment appropriations available in 2015 were higher than in any previous year. Two decisions, permitted under EU budgetary rules and linked to the delay in approval of programmes of the current MFF, contributed to this:

(i) a Council decision to transfer 16,5 billion euro of unused commitments from 2014 to 2015, increasing the MFF ceiling to 163 billion euro;

(ii) a Commission decision to carry over 12,1 billion euro of 2014 unused commitment appropriations, of which 11 billion euro were near completion.

This increase in commitment appropriations for 2015 was expected, as it included those additional commitments linked to Cohesion policy which were supposed to be made in 2014, but have been postponed in line with the applicable legal framework (due to delays in adopting operational programmes under Cohesion). Further increases in commitment appropriations available resulted from assigned revenue (6,8 billion euro) and other amending budgets (0,5 billion euro). Total commitment appropriations available therefore amounted to 181,3 billion euro, equivalent to 1,3 % of EU GNI. Only 2 billion euro of payment appropriations were available for carryover to 2015. The transfer of commitment appropriations therefore increases the pressure on future payment budgets. With most of these commitments being subject to an N + 3 rule, it is normal that these additional commitments do not lead to a related increase in payments. Instead they lead indeed to an increase in the level of outstanding payments RAL (reste à liquider). Outstanding commitments are those commitments that have been made but not cleared by payment or cancelled. These reached 217,7 billion euro (see Figure 2.2). Liabilities and accrued costs are reflected in the balance sheet without necessarily requiring a commitment (for example, for pensions earned by EU civil servants, or the likely cost of decommissioning EU nuclear facilities). Liabilities and accrued costs not covered by a budgetary commitment add a further 121 billion euro, giving a total to be funded from future budgets of 339 billion euro at 31 December 2015 (2014: 305 billion euro). As we have noted in previous years, the Commission does not prepare and publish an annually updated cash flow forecast, spanning a seven to ten year time horizon, covering budgetary ceilings, payment needs, capacity constraints and potential cancellation of commitments. See the Commission’s reply to paragraph 2.40. Evolution of accumulated outstanding commitments (billion euro) Source: Consolidated annual accounts of the European Union — Financial years 2007-2015. THE COURT’S OBSERVATIONS The initial budget for payments in 2015 was higher than in previous years … The EU plans and reports the budget for payments on a cash basis and must balance receipts and payments: the budget cannot be funded by borrowing. The initial EU budget for payments in 2015 was higher than in any previous year. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES … and payments on assigned revenue and appropriations carried over mean that final payments were higher than the MFF ceiling The MFF Regulation adopted in 2013 set a maximum level for payment appropriations for 2015 at 142 billion euro. The initial budget was set just below this ceiling at 141,2 billion euro. The final level of payments was 4 billion euro higher than the level set in the original budget (see paragraph 2.2 and Figure 2.4), and was second only to 2013. For the third successive year the final level of payments was higher than the MFF ceiling for the year. Figure 2.3 shows the gap between the MFF ceiling and final payments over the last seven years. The Court is comparing the voted budget with the level of payments including:

payments made on appropriations generated by assigned revenue (these are on top of the payment appropriations authorised in the 2015 budget and the MFF ceiling) and

payments related to appropriations carried over from previous years (which are to be counted against the MFF payment ceiling of the year, when the appropriations had been authorised by the European Parliament and the Council in the first place).

Final payments compared to the MFF ceiling (billion euro) Source: Consolidated annual accounts of the European Union — Financial years 2009-2015 and MFF Regulations. Payment appropriations and payments made in 2015 (billion euro) Source: Consolidated annual accounts of the European Union — Financial year 2015, Aggregated reports on the implementation of the budget and explanatory notes, tables 5.1 and 5.3. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Parliament and Council approved eight amending budgets. The first was the biggest and dealt with the reprogramming of 16,5 billion euro (paragraph 2.6). The other amending budgets together increased available commitment appropriations by 486 million euro and that available payment appropriations by 85 million euro. Notably they:

re-allocated 1,4 billion euro of commitment appropriations to the EFSI Guarantee Fund, largely from the Connecting Europe Facility (see paragraph 2.35);

added 407 million euro of commitment appropriations to the budget for migration; and

re-allocated 127 million euro of payment appropriations to the migration budget from other headings.

Cost claims for cohesion fell Authorities in the Member States presented a lower level of cost claims for reimbursement for cohesion activities in 2015 than in 2014 (42,2 billion euro in 2015; 2014, 56,6 billion euro). This led to a fall in the level of unpaid cost claims from 23,2 billion euro at the end of 2014 to 10,8 billion euro at the end of 2015. Claims received in the last month of the year nearly halved to 10,5 billion euro (2014, 19,4 billion euro), as did claims received in the last four days of the year (2015, 5,4 billion euro; 2014, 10,1 billion euro). Of the unpaid total of 10,8 billion euro at year end, 2,8 billion euro had remained unpaid since the end of 2014. The EU budget has provided pre-financing to all Member States for both the 2007-2013 and 2014-2020 MFFs. In total uncleared pre-financing amounts to 32,7 billion euro at the end of 2015. Thus the unpaid bills total is not an indication that Member States are overall owed funds from the budget. The Common Provisions Regulation 2014-2020 sets out that the programming period’s pre-financing shall be used for financing of operations carried out under the programming period in question. The period from initial commitment to acceptance of expenditure continues to be long Figure 2.5 provides a breakdown by age of outstanding commitments and pre-financing. The average length of time from commitment to payment is 2,4 years (2,1 years in 2014). In most cases, the first payment on a programme or project (reducing the level of outstanding commitment) consists of pre-financing. Pre-financing remains outstanding until the recipient of EU funds delivers the service or justifies the use of EU funding. The outstanding amount (stock) of commitments is equivalent to nearly two and half times the amount recorded in one year (the flow). Gross outstanding pre-financing is equivalent to just over half a year of payments. The combined average period from commitment to acceptance of expenditure amounts to three years across the EU budget. Use of pre-financing is much more pronounced for indirect management/direct management areas of spending (where the stock of pre-financing is equivalent to 1,6 years payments) than for shared management. For heading 4 the total amount of commitments and of pre-financing outstanding is equivalent to 4,6 years of normal spending. The Commission considers that the reste à liquider (RAL or outstanding commitments) is consistent with the relevant rules for automatic de-commitments. Lengthy delays between initial commitment and final clearing of expenditure is linked to problems of missing documentation at closure and reduces flexibility. As we note in paragraph 2.3 payments in 2015 frequently continued to relate to programmes of previous periods. For cohesion the final date of eligibility is 31 December 2015. Closure documents must be submitted by 31 March 2017. Regulation (EU) No 1083/2006 foresees that a share of 5 % of the total EU allocation is retained until closure of the programme. The co-legislators decided to amend the regulation and extend the N+2 automatic de-commitment deadline for two Member States. This has reduced the pressure on these Member States to submit payment claims within a shorter reference period. Additionally, Croatia’s accession treaty allows an exception to the automatic de-commitment rule. It also provides for N + 3. Finally, a taskforce for better implementation was set up in 2014 to help Member States optimise the use of the 2007-2013 Cohesion policy envelopes. Progress is being monitored by the Commission. Breakdown of the Commission’s outstanding budgetary commitments and pre-financing (million euro) Outstanding budgetary commitments Heading < 2009 2009 2010 2011 2012 2013 2014 2015 Total Years of commitments 1.a Competitiveness 295 672 1209 1541 3885 5942 7844 13066 34455 2,1 1.b Cohesion 1653 38 237 1455 7191 29459 21324 65016 126372 2,6 2 Natural resources 223 62 82 127 213 7231 2140 18112 28191 1,8 3 Security and citizenship 21 39 62 136 277 580 350 1671 3137 1,5 4 Global Europe 938 522 883 1412 3364 4719 5390 7446 24673 3,0 5 Administration 0 0 0 0 0 0 0 295 295 0,0 Total 3130 1333 2473 4671 14931 47931 37049 105606 217123 2,4 Pre-financing Heading < 2009 2009 2010 2011 2012 2013 2014 2015 Total Years of payments 1.a Competitiveness 127 165 422 1382 3353 6117 5343 8636 25545 1,5 1.b Cohesion 12934 9347 689 69 33 75 2042 7523 32711 0,6 2 Natural resources 3443 131 60 78 120 172 393 2397 6795 0,1 3 Security and citizenship 19 27 58 92 192 675 792 1430 3284 1,6 4 Global Europe 236 461 397 817 947 1634 2520 5446 12458 1,6 5 Administration 0 0 0 0 0 0 0 2 2 0,0 Total (gross pre-financing) 16758 10130 1626 2439 4645 8673 11090 25434 80795 0,6 THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Backlogs in the use of European Structural and Investment (ESI) funds are significant By the end of 2015 payments to Member States for the 2007-2013 MFF ESI funds had reached 400,8 billion euro (90 % of the total of 446,2 billion euro for all approved operational programmes). We provide an analysis of the extent to which the Commission has made payments against commitments for each Member State in Figure 2.6. Five Member States account for more than half of the unused commitments of ESI funds that have not led to payment. The five Member States indicated by the Court account for the most significant ESI allocations. In relation to the issue of low absorption levels, the taskforce for better implementation set up by the Commission is actively involved in assisting Member States most concerned by low absorption levels. In some Member States the unclaimed EU contribution, together with required national co-financing, is equivalent to a significant share of total general government expenditure (see Figure 2.7). In seven Member States the total accumulated share that could be claimed from EU funds (taking both the 2007-2013 and 2014-2020 periods) is more than 15 % of annual general government expenditure. A task force for better implementation in Member States was set up in 2014 to improve implementation and budget absorption in the Member States with the lowest absorption and proportionally highest outstanding EU contributions. In order to use all the funds available under the 2007-2013 MFF, authorities in the Member States will need to submit valid claims equalling the sum of all outstanding commitments together with required national co-financing and the amount of pre-financing already provided from the budget. Figure 2.8 shows the relative shares of unused 2007-2013 commitments for each Member State together with the pre-financing paid for the 2014-2020 MFF at the end of 2015. Absorption in % and totals of 2007-2013 MFF ESI funds at 31 December 2015 Croatia acceded to the EU in July 2013. Absorption figures for Croatia are therefore not comparable to those of other Member States. Source: European Court of Auditors calculation based on Commission accounting data. Unused commitments of ESI funds at 31 December 2015 as a percentage of 2015 general government expenditure Source: European Court of Auditors based on information from the Commission. Eurostat data on general government expenditure for 2015, April 2016. 2007-2013 ESI Funds unused commitments and 2014-2020 pre-financing at the end of 2015 (million euro) Source: European Court of Auditors based on information from the Commission. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Financial instruments form a growing part of the budget … Financial instruments are principally composed of loans, equity instruments, guarantees and risk sharing instruments. Financial instruments under indirect management In the 2014-2020 MFF, financial instruments under indirect management have gained in importance and form a growing part of disbursements from the EU budget. The EIB Group manages almost all of the financial instruments under indirect management. This includes the Guarantee Fund for External Actions (see Figure 2.9), which represents 32 % of the total assets held in financial instruments. Only a few very small instruments (assets of 74 million euro), and some technical assistance facilities set up prior to 2014, are managed by other international financial institutions. For the 2014-2020 MFF the Commission has signed only two delegation agreements with bodies outside the EIB Group. The Guarantee Fund for External Action is not a financial instrument in the sense of the Financial Regulation. It pre-existed the Financial Regulation and is based on a lex specialis, Council Regulation (EC, Euratom) No 480/2009 of 25 May 2009 establishing a Guarantee Fund for external actions. Total assets of 6,5 billion euro (1) at 31 December 2015 in financial instruments under indirect management by type of instrument See the table on financial instruments in the FSDA of the accounts. Source: Consolidated annual accounts of the European Union — Financial year 2015 (FSDA section on financial instruments) and compilation of data from financial instruments’ 2015 reporting packages. THE COURT’S OBSERVATIONS The Commission is closing financial instruments from the previous MFFs or merging them into new ones. In the 2014-2020 MFF:

expected payments into financial instruments under indirect management almost double to 7,4 billion euro (2007-2013 MFF: 3,8 billion euro);

the EU budget provides an additional guarantee of up to 16 billion euro to the newly created EFSI (see paragraph 2.33);

the EU Initiative for Small and Medium-sized Enterprises (SME Initiative) (originally designed as a crisis-response instrument) combines EU budget resources available under COSME and Horizon 2020 (175 million euro each), and ERDF-EAFRD (8,5 billion euro) with EIB Group resources.

At the end of 2015, cumulative payments into the new instruments amounted to 1,3 billion euro (2014: 0,7 billion euro) as shown in the Figure 2.10. Of this:

guarantee instruments amount to 1,1 billion euro (88 %) of cumulative payments;

equity instruments amount to 0,2 billion euro (12 %) of cumulative payments.

2014-2020 MFF financial instruments under indirect management Source: SWD(2015) 206 final, Delegation Agreements and ABAC Workflow. (million euro) Name of the instrument Type of instrument Indicative budget 2014-2020 Cumulative commitments Cumulative payments 2014 2015 2014 2015 HORIZON 2020 — InnovFin Loan and Guarantee Service for RDI — EIB Guarantee 1060 483 646 483 646 of which earmarked revenue re-allocated from RSFF 375 430 374 395 HORIZON 2020 — InnovFin SMEs & Small Midcaps R&I Loans Service — EIF Guarantee 1060 169 306 169 304 HORIZON 2020 — InnovFin Equity Facility for R&I — EIF Equity 430 0 125 0 110 CEF Debt Instrument — EIB Guarantee 2557 128 247 0 0 CEF Equity Instrument Equity 0 10 0 0 COSME Equity Facility for Growth — EIF Equity 633 52 102 33 41 COSME Loan Guarantee Facility — EIF Guarantee 686 89 238 42 131 Student Loan Guarantee Facility (Erasmus +) — EIF Guarantee 517 28 61 0 20 Employment and Social Innovation (EaSI) — EIF Guarantee 193 28 60 0 10 Cultural and Creative Sector Guarantee Facility Guarantee 123 0 0 0 0 Private Finance for Energy Efficiency Instrument (PF4EE) — EIB Guarantee 80 30 50 6 12 Natural Capital Financing Facility (NCFF) — EIB Guarantee 60 10 30 3 12 Total 7399 1017 1875 735 1286 THE COURT’S OBSERVATIONS … and of the balance sheet … Total assets held in financial instruments under indirect management increased by 650 million euro (11 %) on the previous year (see Figure 2.11). Key elements of this increase are:

a 251 million euro increase in investments in debt securities and fixed income securities resulting from transfers from the EU budget to fiduciary accounts;

a 177 million euro increase in investments in euro-denominated bonds by the Guarantee Fund for External Actions, to which 144 million euro had been transfered from the EU budget.

Total assets of financial instruments under indirect management by type of asset (billion euro) Source: Consolidated annual accounts of the European Union — Financial years 2014 and 2015. THE COURT’S OBSERVATIONS Cash represents 20 % of financial instruments’ total assets, and remained stable at 1,3 billion euro at the end of 2015 (2014: 1,3 billion euro) (see Figure 2.12). Cash and cash equivalents belonging to financial instruments (1) See Note 2.8 of the 2015 accounts. Source: Compilation of data from financial instruments’ 2015 reporting packages. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES During 2015, 211 million euro were returned from the financial instruments’ trust accounts managed by the entrusted entities: 126 million euro to the general budget and 85 million euro as internal assigned revenue. Of this:

134 million euro represent revenues and repayments;

59 million euro represent returns to the general EU budget of unused funds held on the trust accounts of financial instruments and technical assistance facilities at closure and 18 million euro represent cancelled European Neighbourhood and Partnership Instruments (ENPI) operations.

For the financial year 2015, total administrative and management fees for financial instruments amount to 63 million euro (2014: 50 million euro). For the new instruments cumulative fees for 2014 and 2015 amount to 72 million euro (5,77 % of total assets held). … and unutilised amounts in financial instruments under shared management remain relatively high According to the latest available report on financial instruments under shared management, describing the situation as at the end of 2014, 6,8 billion euro of unutilised amounts remained in these instruments. Of this 80 % was concentrated in five Member States. Italy (3 billion euro) constituted 45 % of the total. During our examination of the balance sheet we asked the Commission to provide additional analysis on the use of these instruments in Italy up to 31 December 2015. This shows that 1,7 billion euro remained unutilised at 31 December 2015 for Italy. The Commission requested additional information from Italy by 31 December 2015. This information shows that 45 % for Italy remained unutilised as at 31 December 2015. The data obtained, checked against the bookings made in the 2015 Commission accounts, have been transmitted to the Court. Following the analysis, it appeared that there is no need to change the current methodology. However, implementation will be reassessed for the 2016 exercise. The first operations were recorded for the European Fund for Strategic Investments (EFSI) … The Commission (representing the EU) and the EIB signed the EFSI Agreement on 22 July 2015. EFSI is part of the Investment Plan for Europe announced in 2014 by the President of the European Commission. It aims to support investments in areas such as infrastructure, education, research and innovation, as well as risk finance for small businesses. The EU provides a budget guarantee of up to 16 billion euro (of which 8 billion euro will be transferred gradually to a specific fund between 2016 and 2022) and the EIB contributes 5 billion euro from its own resources to a risk reserve. The Commission and the EIB expect this to support EIB/EIF financing of up to 61 billion euro, with the objective of leveraging 315 billion euro of investments. The EU guarantees Infrastructure and Innovation Window (IIW) operations, implemented by the EIB, and Small and medium-sized enterprises Window (SMEW) operations, implemented by the EIF. Each of these has a debt portfolio (EU guarantee of 11 billion euro for the IIW and 1250 million euro for the SMEW) and an equity portfolio (EU guarantee of 2500 million euro for the IIW and 1250 million euro for the SMEW). The Commission will transfer funds from the guarantee to the EIB on receipt of the EIB’s request. To meet such claims the Commission set up a guarantee fund (EFSI guarantee fund). The EU budget will finance this fund over the period to 2020 by reducing planned spending on the Connecting Europe Facility (CEF) by 2,8 billion euro and on Horizon 2020 by 2,2 billion euro and by using 3 billion euro from unallocated margins under the MFF expenditure ceiling. In December 2015 the Commission committed 1,4 billion euro for payments into the fund. It made no payments in 2015. The funds will be paid to the EIB if required to honour guarantee calls in line with the EFSI Regulation under the conditions specified in the EFSI agreement. The EIB signed eight IIW debt portfolio operations (totalling 1092,7 million euro) of which it disbursed 193,8 million euro in 2015. The EIB signed two IIW equity portfolio operations (totalling 125 million euro in 2015), half of which is guaranteed by the EU. Of the amounts covered by the EU guarantee, the EIB disbursed only 7,6 million euro in 2015. The total EU guarantee on disbursed amounts and accrued interest on this portfolio amounted to 202 million euro at year end. Disbursements increase gradually when building up a portfolio of operations. The EFSI portfolio is currently being expanded with a considerable increase in size foreseen by the end of 2016. For the SME window, during 2015 the EIF used the EU guarantee only to guarantee operations under existing SME financial instruments. … while launch of the Connecting Europe Facility (CEF) was delayed The Commission and the EIB signed the CEF debt instrument agreement on 22 July 2015, the same day on which they signed the EFSI agreement. Merger of the LGTT and PBI portfolios was postponed until 2016. The Commission paid no contribution to the instrument in 2015 and the EIB signed no operations. CONCLUSIONS AND RECOMMENDATIONS The conclusions for 2015 The proliferation of financial mechanisms which are not directly funded by the EU budget nor audited by the Court poses risks both for accountability and the coordination of EU policies and operations (paragraph 2.4). The Commission made a high level of commitments in 2015, in part on appropriations brought forward from 2014. Outstanding commitments returned to previous levels (paragraphs 2.5 to 2.9). Amounts to be funded from future budgets rose to 339 billion euro. The Commission has not produced a cash flow forecast covering the next seven to ten years (paragraph 2.10). Such a forecast would enable stakeholders to anticipate future payment requirements and budgetary priorities and constructively plan ahead. The MFF is the tool ensuring medium- to long-term stability and predictability of future payment requirements and budgetary priorities. Therefore long-term forecasts can only be established taking into consideration existing financial frameworks. The payment implications of future financial frameworks which have not yet been agreed cannot be predicted with precision. The forecast for the post-2020 period will be presented in 2017 (as foreseen in point 9 of the Inter-institutional Agreement on budgetary discipline, on cooperation in budgetary matters and on sound financial management) and it will be an integral part of the Commission proposals for the post-2020 MFF (to be presented before 1 January 2018 based on Article 25 of the MFF Regulation). The Commission already committed itself to presenting a medium-term payment forecast (in the framework of the Mid-term review of the financial framework) which will analyse the sustainability of the payment ceilings and evolution of RAL and de-commitments until 2020. Cost claims submitted for cohesion during 2015 fell. This led to a reduction in cost claims awaiting reimbursement, but did not reduce the level of payments made (see paragraphs 2.14 and 2.15). Continuing lengthy delays between initial commitments and final clearings of expenditure means that patterns of spending are slow to respond to changing budgetary priorities and increases the risk that documentation is unavailable at closure (see paragraphs 2.16 and 2.17. The delay between commitments and payments are just reflecting the duration of ongoing projects. Some projects are large infrastructure projects in the Union or projects in rather difficult third countries and an average delay of 2,4 years should not be seen as particularly abnormal. The legal framework obliges the beneficiaries to keep the relevant documentation for a certain period after the closure. Outstanding commitments (together with outstanding pre-financing) in ESI Funds of the 2007-2013 MFF for some Member States remain high and frequently exceed 15 % of general government expenditure (paragraphs 2.19 to 2.21). Increasing amounts are held in financial instruments under indirect management, for such the EIB Group remains the preferred financial institution for the management (paragraph 2.23). Unused amounts of the financial instruments under shared management remain high and are concentrated in some Member States (paragraph 2.31). See reply to paragraph 2.31. Establishment of EFSI affected the launch of the CEF programme. Although it affects the assumptions the Commission made in its ex-ante assessment of the use of financial instruments under CEF, this was not re-evaluated. EFSI will also impact upon the use made of some other financial instruments (see paragraphs 2.32 to 2.38). The adoption of EFSI did not affect the overall objectives of CEF, but created an additional mechanism for the EU financial assistance, i.a. to infrastructural projects, in particular via debt finance. To take into account the launch of EFSI, the CEF DI Steering Committee in its first meeting defined the main principles for providing CEF and EFSI support, with a view to enhance their complementarity. Moreover, an addendum to the CEF ex ante assessment is under preparation to reflect (i) the latest analysis on the potential to launch a CEF Equity Instrument in the CEF sectors, and (ii) the market developments post-EFSI launch. This addendum is to be understood as complementing the previous ex ante assessment, and not as a revision. The coherence of CEF debt instrument with other EU initiatives, including EFSI, will be assessed in the mid-term review of the CEF. Recommendations We recommend that:

Recommendation 1: the Commission should act to reduce outstanding commitments by, for example, quicker decommitment, where appropriate, faster closure of the 2007-2013 programmes, where appropriate, wider use of net correction in cohesion, reducing cash held by fiduciaries and the compilation of payment plans and forecasts in those areas where outstanding commitments and other obligations are significant (see paragraphs 2.5 to 2.9, 2.19 and 2.28);

The Commission accepts the recommendation. Any measures taken by the Commission vis-à-vis the Member States, including efforts to reduce outstanding commitments and make wider use of net corrections in cohesion must obviously comply with the regulatory framework

Recommendation 2: the Commission should prepare and publish an annually updated cash flow forecast, spanning a seven to ten year time horizon, covering budgetary ceilings, payment needs, capacity constraints and potential decommitments (paragraph 2.10);

The Commission accepts the recommendation. The forecast for the post-2020 period will be presented in 2017 (as foreseen in point 9 of the Inter-institutional Agreement on budgetary discipline, on cooperation in budgetary matters and on sound financial management) and it will be an integral part of the Commission proposals for the post-2020 MFF (to be presented before 1 January 2018 based on Article 25 of the MFF Regulation). In addition the Commission will present a mid-term payment forecast assessing the sustainability of the current payment ceilings until the end of the current MFF in the framework of the Mid-term review of the MFF.

Recommendation 3: the Commission should consider in its budgetary and financial management the capacity constraints in certain Member States in order to ensure funds are used efficiently (see paragraphs 2.19 to 2.21);

The Commission accepts the recommendation within the framework set out by the MFF and the relevant legislation. The Commission has already put in place a Task Force for Better Implementation in order to improve the implementation of ESI funds in certain Member States, while ensuring full respect of legality and regularity of expenditure (see the Commission’s reply to observations 2.19 and 2.20). The Commission intends to ensure that full benefit is drawn from the work of the Task Force and will focus on delivering a proactive and targeted approach to Member States in order to support the implementation of the Cohesion policy funding in 2014-2020.

Recommendation 4: the Commission should recover unused cash balances in financial instruments under shared management (see paragraph 2.31) and remaining unused funds in indirect management financial instruments from previous MFFs for which the eligibility period has expired;

The Commission accepts the recommendation. For the indirectly managed financial instruments, a procedure for return and de-commitments of funds to the EU covering the transfer of GIF revenues and repayments to H2020 and COSME was finalised by the Commission with the EIF in November 2015 and execution is under way. It will be accompanied by monitoring visits to the EIF. For the shared management, any unused financial instruments allocations will be deducted at closure in accordance with the legal provisions for shared management programmes.

Recommendation 5: the Commission should re-evaluate the ex-ante assessment for the CEF debt instrument in the light of creation of the EFSI, and consider the impact of EFSI on other EU programmes and financial instruments.

The Commission partially accepts the recommendation. The coherence of CEF debt instrument with other EU initiatives, including EFSI, will be assessed in the mid-term review of the CEF. An addendum to the CEF ex ante assessment is under preparation to reflect (i) the latest analysis on the potential to launch a CEF Equity Instrument in the CEF sectors, and (ii) the market developments post-EFSI launch. This addendum is to be understood as complementing the previous ex ante assessment, and not as a revision The additionality of CEF financial instruments towards EFSI is based on a principle that CEF financing focuses on innovative, demonstrator and pilot projects and initiatives eligible under the relevant regulations in the CEF sectors, while the scope and eligibility criteria of EFSI support are broader.

CHAPTER 3 Getting results from the EU budget

Txt. 1. THE COURT’S OBSERVATIONS INTRODUCTION The subject of this chapter is performance (see Figure 3.1). It continues the theme — which has been a feature of our annual reports since 2010 — of including a chapter assessing different aspects of the systems put in place by the Commission to ensure that EU funds are spent well, and not just in line with relevant legislation. Since 2010, we have analysed the Commission’s self-assessment on performance as stated in the annual activity reports (AARs) of a selection of the Commission’s directors-general, and included an examination of some of the themes developed in that year’s special reports on performance audit. Other elements of performance that we have examined are, the management plans of selected directorates-general (DGs) and the reporting by the Commission to the European Parliament and the Council including the first few editions of the Commission’s evaluation reports. Performance in the EU Performance in the EU is assessed on the basis of the sound financial management principles (economy, efficiency and effectiveness), and covers: (a) inputs — financial, human, material, organisational or regulatory means needed for the implementation of the programme; (b) outputs — the deliverables of the programme; (c) results — the immediate effects of the programme on direct addressees or recipients; (d) impacts — long-term changes in society that are, at least partly, attributable to the EU’s action. Txt. 2. In the equivalent chapter of last year’s annual report, we took a particular look at the EU’s long-term strategy, Europe 2020. We examined the EU’s high-level monitoring and reporting of the strategy and, for the European Structural and Investment Funds (the ESI funds), how the strategy feeds through into the partnership agreements and funding programmes agreed between the Commission and Member States. This year we continue that theme, and focus our examination of the Commission’s performance through Horizon 2020, the Commission’s research and innovation programme and, as will be seen below, a very significant item in the EU budget. The chapter therefore consists of three parts. Horizon 2020 performance is the subject of the first part. The remaining two parts deal with subjects we have examined in previous years: the extent to which performance is covered in the main annual planning and reporting documents prepared by the Commission’s directors-general and some of the main themes arising from our 2015 special reports including the follow-up of a sample of recommendations from previous special reports and, for the first time, responses from the Member States to a survey that we conducted on their follow up of our recommendations. In the 2014 annual report, we reported that, based on information supplied by the Commission, the mid-term review of Europe 2020, which should have been available to the Council in March 2015, had been delayed until early 2016. As of June 2016, the Commission informed us that the mid-term review had been finalised and that there would now be no formal mid-term review concluding document. As a result, we are not in a position to report in this chapter on how this review process was carried out. We may, however, return to this issue in future reports. PART 1 — HORIZON 2020 (2014-2020) There are three elements to our analysis of Horizon 2020. First we set out the main features of the programme and examine the strategic framework in which it operates, in particular its links to other EU strategies. The second element considers the advantages and potential risks involved in bringing research and innovation activity together. Finally we analyse the extent to which there is a focus on performance within Horizon 2020. Section 1 — Links between Europe 2020, the new Commission priorities and Horizon 2020 do not always facilitate a focus on performance and need to be clarified Europe 2020 and Horizon 2020 Europe 2020 is the European Union’s 10-year jobs and growth strategy. It was launched in 2010 to create the conditions for smart, sustainable and inclusive growth. Responsibility for delivering the Europe 2020 strategy is shared between the EU and its 28 Member States. Thus the strategy’s five headline targets have an associated national target in each Member State, reflecting the different circumstances of each. Horizon 2020 has a budget of around 75 billion euro for 2014-2020. It is the EU’s programme for supporting research and innovation, with the general objective to contribute to building a society and an economy based on knowledge and innovation across the Union. It supports the Europe 2020 strategy and the EU objective of creating a European Research Area (ERA). It brings together previously separate spending programmes for research and innovation. Funding is provided mainly in form of grants, although access to other forms of finance, such as financial instruments, is also available. The Commission manages the majority of expenditure under Horizon 2020, either itself or through agencies and other bodies. Beneficiaries of Horizon 2020 are individuals, teams, universities, research institutes, SMEs and larger companies, from within and outside the EU. Many funded projects are collaborative. Figure 3.2 shows the structure of Horizon 2020. The general objective is translated into three priorities (1. Excellent science, 2. Industrial leadership and 3. Societal challenges). These priorities are pursued through 14 specific objectives, complemented by broad lines of activities which provide their scientific and technological content. The general objective is also pursued through two specific objectives (Spreading excellence and widening participation and Science with and for society), while two additional organisations, the JRC and the EIT also contribute to the general objective and the priorities by following their own specific objectives. Horizon 2020 also includes additional research measures, such as Fast-Track to Innovation to promote close-to-the-market innovation activities. The structure of Horizon 2020 Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES There are high-level links between Horizon 2020 and Europe 2020 but need to be enhanced to ensure that they facilitate a greater focus on performance Under the legislation, Horizon 2020 has three performance indicators for assessing progress against its general objective: Pt_a_. 2. the research and development intensity indicator according to which investment should account for at least 3 % of GDP by 2020 — replicating one of Europe 2020’s five headline targets; Pt_b_. 2. an innovation output indicator; Pt_c_. 2. the share of researchers in the active population. This indicator has no direct link with Europe 2020 headline targets. Txt. 3. The 3 % indicator measures the input of funds in research and development. It was not designed to measure output, results or impact. Horizon 2020 adds for the first time in legislation the requirement that private — as opposed to public — investment in research and development should account for two-thirds of the total. Against these targets, in 2014 total such investment across Europe was just over 2 % of GDP, with the private sector accounting for around 64 % of this. Comparable figures for the USA were 2,8 % in 2012, 3,5 % for Japan and 4,2 % for South Korea (for both countries in 2013). Whilst the 3 % R&D intensity indicator measures the input of funds in research and development, it also reflects the effectiveness of policy measures taken to stimulate a high quality science base in the Member States and to create better framework conditions for research and innovation in the private sector. Horizon 2020 spending is one relatively small contribution to the target. It was always expected that Member States would provide the bulk of funds in this area. In 2014, the first year of implementation, Horizon 2020 spending by the EU institutions represented just under 10 % of total public budgets for R&D. We noted in last year’s annual report that meeting all national targets for research and development would lead to a total investment of 2,6 % of GDP by 2020, some way below the 3 % target. In the area of research, as well as Horizon 2020, efforts are made to coordinate and improve Member States’ investment in research and innovation, and to develop framework conditions that allow private sector innovation to thrive (open innovation). This is crucial to making significant progress in closing the innovation divide between the European Union and its key competitors. Increasing impact and excellent science through openness (open science) and a digital single market, embedding digital into the grand societal challenges, implementing the research, innovation and competitiveness dimension of the Energy Union, together with a forward-looking climate-change policy, and translating Europe’s strengths in science and technology into a leading global voice (open to the world) are also key objectives of research policy. Research and innovation investments from the cohesion policy programmes for 2007-2013 also contributed to the progress towards the 3 % target. This contribution is continued for the 2014-2020 programming period thanks to the thematic concentration of ERDF and will strengthen its impact and leverage thanks to the Smart Specialisation ex ante conditionality for research and innovation. The contribution of Horizon 2020 to the achievement of the Europe 2020 strategy is reinforced by the interaction with other actions at EU and national level. The second Horizon 2020 performance indicator, on innovation, is a composite output indicator. It was initially developed for the Europe 2020 flagship initiative Innovation Union. It captures information on four elements: patents, employment in knowledge-intensive activities, competitiveness of knowledge-intensive goods and services, and employment in fast-growing firms of innovative sectors. The Composite Indicators Research Group (financed by the EU) has pointed to the fact that such indicators, based on sub-indicators with no common meaningful unit of measurement and with no obvious method for weighting, have pros and cons. In particular, while they illustrate the big picture, there is a risk that in isolation they can lead to simplistic policy conclusions; they should therefore be used in combination with the sub-indicators. The Commission reported on this indicator and the sub-indicators in March 2016. The innovation output indicator has the advantage that it focuses attention on establishing the right conditions for innovation and that it provides a measure for assessing progress in innovation at the level of the EU economy. However, it also has a number of shortcomings which limit its usefulness for monitoring the Horizon 2020 contribution towards the Europe 2020 strategy objectives: 3.15. (8) the Commission has not yet set a quantified target to be reached by a defined date; (9) In order for a quantitative target to impact policymaking and stimulate progress, the target set must be consensual, justified on the basis of evidence and benchmarking, and easy to understand and communicate. This is more difficult to achieve in the case of composite indicators, such as the Innovation Output Indicator, as the summary scores in which they result are by nature abstract. Therefore, composite indicators typically do not come with target levels. Another example of this is the Summary Index of the Innovation Union Scoreboard, which also does not have a target level, but still clearly has an impact on policymaking of Member States. Furthermore, the available time series for some of the component indicators are still too short to establish trend lines and identify a meaningful EU target level (the component on high-growth enterprise was based on a new Eurostat data collection). In addition, setting an EU target level would require benchmarking against the performance non-EU countries, something not possible at the time the indicator was launched. (10) the Commission has not translated it into national targets to reflect the individual situation of each Member State — as recommended by experts in the field for indicators measuring innovation. (11) The Research intensity target is set by Member State, and the Innovation Output Indicator complements it by providing the policy picture on innovation output. These different elements need to be seen together. Txt. 4. The specific objectives set by Horizon 2020 are the next level in the hierarchy of objectives where links are established to the Europe 2020 strategy. The largest element (38 %) of the Horizon 2020 budget is allocated to the priority Societal challenges which is designed to address policy priorities in the Europe 2020 strategy. In many cases, the description of the specific objectives refers to the flagship initiatives they are designed to support. However, although they are a useful indication of the generic relationship between Horizon 2020 and the Europe 2020 strategy, they are not further translated at lower levels into more specific and operational links that could serve as the basis for monitoring and reporting the contribution of Horizon 2020 to the individual flagship initiatives and thus the Europe 2020 strategy. For the preparation of the budget, the contribution of Horizon 2020 to each flagship initiative is provided in the Horizon 2020 programme statement on an annual basis, following a methodology developed with the launch of the programme. The links between the Commission’s 10 political priorities and Europe 2020/Horizon 2020 need further clarification In 2013, the regulation establishing Horizon 2020 set priorities, general and specific objectives for the period 2014-2020. These priorities and objectives were designed to be aligned with, and to contribute to, the Europe 2020 strategy. Towards the end of 2014, the current Commission set 10 priorities for the Commission as a whole for 2014-2019. While these priorities highlight some new areas (such as a free trade agreement with the United States), they also overlap significantly, as one would expect, with the existing Europe 2020 strategy, including its headline targets and flagship initiatives. For example, employment and climate change are issues which feature heavily in both. Despite the overlaps and complementarities, it has been difficult to establish the extent to which the 10 political priorities form a distinct new approach of the Commission or are intended as a new way for the Commission to put Europe 2020 into operation for its work. The Europe 2020 strategy is the growth strategy for a decade for the EU and at EU level. In line with the Europe 2020 strategy, the 10 political priorities of the Commission provide a clear and focused agenda/strategy for the activities of the Commission towards Europe 2020 and they are entirely consistent with the overall Europe 2020 goals. At the same time, they provide greater clarity and focus in the Commission’s overall political strategy, taking into account recent developments. For example, since the establishment of the Europe 2020 strategy the EU has seen sovereign debt crises, increasing public sector deficits, war in Ukraine and Syria, a massive flow of migrants towards it, and the COP 21 climate agreement. These 10 political priorities have also been presented to the European Parliament, the European Council and the Council in the course of the appointment of the new Commission. The priorities take over, as targets and indicators, the targets and indicators of the Europe 2020 strategy. It is normal for strategic priorities to alter, as circumstances change. However, as pointed out by the Commission’s internal audit service, although there is common ground between the two strategies, the Commission has, so far, not mapped the correspondence between them. In our view, since the relationship between the two strategies has not been clarified, they coexist and are used for different purposes: The Commission considers that its priorities provide a clear and focused policy agenda for the Commission that is consistent with the Europe 2020 strategy for the European Union as a whole. As such, they are naturally used for different purposes. However, this does not mean that they are in any way inconsistent or incoherent, they are entirely complementary. Please refer also to the Commission reply to paragraph 3.18. (12) The allocation of the EU budget to spending programmes (such as Horizon 2020) is based on, among other things, their contribution to the Europe 2020 headline targets and flagship initiatives. The monitoring and evaluation of spending programmes also builds on objectives designed to reflect the Europe 2020 strategy. (13) In relation to Horizon 2020 this was implemented in the first work programmes, for 2014-2015. Over the remainder of the programme the work programmes, and the budget for particular elements of the work programme, will be increasingly influenced by the political priorities set out by the current Commission in response to the emerging political, economic and technological context. (14) The Commission’s annual work programme now focuses on the 10 political priorities. Starting from 2016, the Commission has also introduced a new method of strategic planning, which requires DGs to prepare 5-year strategic plans, also based on the 10 political priorities. The 5-year strategic plans will become the starting point for drafting annual management plans (MPs) for carrying out the DGs’ tasks (including the implementation of spending programmes such as Horizon 2020). Since the reporting in annual activity reports is strongly linked to the MPs, annual activity reports from 2016 onwards can also be expected to reflect primarily the 10 political priorities. (15) The Commission’s annual work programme includes the actions that the Commission will take to help deliver on the political priorities and contribute to the achievement of the Europe 2020 strategy. The new multiannual Strategic Plans developed for each Commission department are built around the 10 political priorities, expressed as general objectives. As part of this exercise, the DGs were instructed to explain how their activities contribute to the Europe 2020 strategy. To facilitate this, the headline targets of the Europe 2020 strategy appear as impact indicators linked to the relevant general objectives. The accompanying annual management plans set out the planned outputs for each department and how these will contribute to the objectives. Commission departments will report on their achievements through their annual activity reports, which will reflect the structure of the strategic and management plans. However, the Commission will continue to report on progress towards the Europe 2020 targets as part of the European Semester process and on the contribution of the budget to Europe 2020 in the annual management and performance report. Txt. 5. For Horizon 2020, this leads to a situation where the legal framework and the budget allocation for the programme reflect the Europe 2020 strategy, while those parts of the Commission implementing Horizon 2020 have refocused strategic planning and management arrangements on the 10 political priorities. For example, the key priorities to be developed in the Horizon 2020 work programmes 2016-2017 are based on seven out of the 10 policy areas on the agenda of the new Commission. Under these circumstances, it is unclear what strategic objectives Horizon 2020 should be targeting. DG RTD identified four political priorities where the DG can make a contribution. The pursuit of these priorities, however, will require a refocusing of effort against the original targeting as set out in the Horizon 2020 legislation. The Commission considers that there is no lack of clarity about which strategic objectives Horizon 2020 should be targeting. These objectives are the Europe 2020 objectives, as set out in the legal base. The 10 political priorities are used to consider how the Commission can best assist in reaching the Europe 2020 priorities. For example, the delivery of the digital single market (DSM) strategy will lead to updated priorities in the work programme 2017-2018 of the ICT part of Horizon 2020 in line with both Europe 2020 and the digital single market Commission priority, within the flexibility available in Horizon 2020. The respective set of measures and initiatives of 18 April 2016 (DSM package) already demonstrates this alignment of Horizon 2020 and the DSM priority. In last year’s annual report we pointed to the weak alignment between the Europe 2020 and the EU’s seven year budgetary cycle. The same complication may affect the 10 political priorities, as the mandate of the current Commission ends in 2019, a year before the end date of Europe 2020 and Horizon 2020. The 2007-2013 MFF period may not have fully coincided with the period covered by the Europe 2020 strategy. However, the Commission maintains its position formulated in reply to paragraphs 3.7-3.11, 3.87 of the Court’s annual report for 2014, that this does not imply a lack of alignment with the overall strategy. Moreover, the 2014-2020 MFF period falls entirely within, and has the same end date, as the current Europe 2020 strategy period. The 2014-2020 MFF has been designed with a view to concentrate on delivering on the Europe 2020 strategy. The Commission is monitoring and reporting on the contribution of the EU budget to the EU’s overall objectives and to the Europe 2020 strategy in particular. The current Multiannual Financial Framework, Horizon 2020 and the Europe 2020 strategy are fully aligned and have a common end date of 2020. These frameworks are not affected by the expiry of the mandate of the current Commission in 2019. Section 2 — Synergies and complementarities with ESI funds and national programmes Limited information on synergies with ESI funds One of the main objectives of the Europe 2020 strategy and its Innovation Union is the completion of the ERA in which researchers, scientific knowledge and technology circulate freely. The EU budget contributes to the ERA objective through Horizon 2020 and the ESI funds. In this context, and in line with the legislation, the Commission envisaged a system for encouraging and tracking synergies between Horizon 2020 and the ESI funds. The classification of Horizon 2020 projects according to ESI thematic objectives, together with the introduction of specific references to potential synergies at their planning and reporting stages would provide the basis for such a system. However, in the first Horizon 2020 monitoring report, information on the use and impact of synergies between Horizon 2020 and the ESI funds is limited. The Commission plans to analyse the impact and synergies between Horizon 2020 and the ESIF in the context of the Horizon 2020 interim evaluation. The Commission has created the regulatory conditions that can enable synergies between the funds by increased coordination and harmonisation of the relevant legal frameworks. However, the implementation of synergies needs a continued effort both on the side of the bodies in charge of Horizon 2020 and of the authorities of the Member States and regions. Significant efforts were undertaken in Member States and regions, in particular in the framework of the development of smart specialisation strategies to develop policy measures that allow to implement the different synergy types identified by the Commission’s guidance and to focus R&I efforts on fields where the Member State or region has the biggest chances for excellence in terms of science and business competitiveness. These fields also offer the biggest chance to be successful in Horizon 2020 calls. Also on the side of Horizon 2020 management, extensive action is being undertaken to promote and develop synergies, for example through the set-up of memoranda of understanding between joint undertakings or the seal of excellence initiative, launched at the initiative by Commissioners Moedas and Creţu in October 2015. A number of regions and Member States are already involved in this initiative. The Commission also implements the Stairway to Excellence initiative helping the EU-13 Member States to improve their access to Horizon 2020. Examples of complementarities between national and EU research programmes The Horizon 2020 legislation recognises the importance of building synergies and complementarities between national and European research and innovation programmes. Figure 3.3 below includes key messages from our cooperation with the Supreme Audit Institutions (SAIs) of Bulgaria and Portugal: while there are some areas in which national and EU research programmes are complementary, our partners also found some problems at the national level. Examples of complementarities with national programmes including some identified weaknesses at national level Examples of complementarities with national programmes including some identified weaknesses at national level The relation between the EU priorities in the context of the Europe 2020 strategy for smart, sustainable and inclusive growth, including Horizon 2020, and the national priorities of Bulgaria, is articulated through eight priorities set in the National Development Programme Bulgaria 2020 (NDP). Three of these priorities are related to science and innovation: (16) Improving the access to and enhancing the quality of education and training and the quality characteristics of the workforce. (17) Support for innovation and investment activities to increase the competitiveness of the economy. (18) Energy security and increased resource efficiency, particularly for SMEs. Txt. 6. In addition to the NDP, which is an integrated document for the socio-economic development of Bulgaria up to 2020, the objectives under Horizon 2020 are also synchronised and broken down into operational objectives and plans in: (19) the National Strategy of Scientific Research (NSSR) to 2020, which is designed to support scientific development in Bulgaria and make it into a factor for the development of an economy based on knowledge and innovation activities, (20) the Innovation Strategy for Smart Specialization (ISSS), that has the strategic objective for Bulgaria to move from the group of modest innovators to the group of moderate innovators by 2020. Txt. 7. The national strategies set a national objective to increase the share of expenditure on research and development activity to 1,5 % of GDP by 2020. To achieve the national objective measures and activities are envisaged in the Action Plans implementing the national strategic documents. However, the BNAO identified that neither quantitative indicators for results, nor target values per year for activities related to Horizon 2020 are included in the action plans for the NDP. Neither were quantitative indicators per activity included in the action plan for the NSSR. The BNAO has also identified some issues regarding coordination and interaction between all the participants engaged in Horizon 2020 at the national level, and that improving the interaction between all the participants would stimulate Bulgaria’s participation in the programme. The Commission supports Bulgaria actively both via the Stairway to Excellence pilot (see above) and the new lagging regions initiative in order to improve its innovation eco-system and participation in Horizon 2020 via the implementation and evolution of the Bulgarian smart specialisation strategy. The European Commission presented in October 2015 key policy recommendations to help modernise the research and innovation sectors in Bulgaria. This was the first time that an EU Member State made voluntary use of the Horizon 2020 Policy Support Facility (PSF), a new instrument that gives governments practical support to identify, implement and evaluate those reforms needed to enhance the quality of their public research and innovation structures. The main national strategic document in the area of research and innovation is the Research and Innovation Strategy for Smart Specialization (RISSS) that has a 2014-2020 timeframe, matching the Horizon 2020 timeframe and Europe 2020 end date. One of its chapters includes links between Europe 2020 and the Portugal 2020 partnership agreement, signed between Portugal and the Commission, in the area of R&D and innovation, mobilizing funds from both cohesion policies and Horizon 2020. The Portuguese Tribunal de Contas noted that the RISSS included two scenarios for the evolution of the indicator on the proportion of GDP invested in research and development: a conservative one of 2 % achieved by 2020, and a change one with 2,7 % by 2020. It also found that, taking account of Europe 2020, Horizon 2020 indicators and the Innovation Union indicators, some issues are not tackled by the national strategy indicators, such as the researchers’ gender, the contribution that is sustainable development and climate change related, and the participation of SMEs. The Tribunal de Contas also noted that there are similar indicators at national and European level, such as those dealing with scientific publications or innovative products or processes, that cannot be compared, for instance because different denominators are used. Section 3 — The mechanics of Horizon 2020 do not consistently drive a focus on performance Measuring performance in the area of research is inherently difficult. The results and impacts of research by their nature are difficult to predict, generally rather long-term and often do not follow a linear path. However, the EU still needs to obtain reliable information on scientific progress, productivity and quality for the research it finances, in order to measure scientific performance for good decision making. This need is recognised in the Horizon 2020 framework regulation which states that effective performance management, including evaluation and monitoring, requires the development of specific performance indicators […] which are relevant to the appropriate hierarchy of objectives and activities. Against this background, this section examines whether: Pt_a_. 3. there is a focus on performance, including a clear and operational link to Europe 2020, in the grant management process of Horizon 2020; Pt_b_. 3. the design of the monitoring and reporting system of Horizon 2020 enables the Commission to monitor and report on performance; and Pt_c_. 3. important basic concepts of performance monitoring and reporting are used in a consistent and correct manner by the Commission. Txt. 8. The grant management processes are not fully designed to focus on performance To assess whether the Horizon 2020 grant management processes included an appropriate focus on performance, particularly in its links to Europe 2020, we examined the design of the Horizon 2020 legal framework, the focus on performance at the level of work programme and calls and, for each of the five work programmes we sampled, one successful proposal by beneficiaries. The legal framework of Horizon 2020 introduces several important elements for performance management but there are some weaknesses in design, particularly as regards the use of performance indicators The objectives and indicators agreed by the legislative authority for Horizon 2020 represent a real improvement on the previous framework programmes. As we have observed in the past, objectives defined in legislation may be at too general and high a level for operational use by management. This is also the case for these high-level Horizon 2020 objectives — even those referred to as specific objectives in the regulation, most of which are not actually formulated as objectives but rather as challenges and/or areas. For example, for the priority Societal challenges two specific objectives are: health, demographic change and well-being; and secure clean and efficient energy. The extent to which these high-level objectives are successfully translated into lower-level objectives by management is examined in the next section. For Horizon 2020, the inclusion of objectives in the legal text, for the first time, was an undoubted step forward. It did lead to an extensive debate with the legislative authority about the objectives of the programme, and the indicators to be monitored. As well as specifying the objectives of the programme, the Horizon 2020 legislation also includes three performance indicators for assessing progress against the general objective, set out in paragraph 3.10. At the next level down of the legal framework, the specific programme introduces further elements important for performance management. It lists 18 key indicators for assessing results and impacts of the specific objectives of the Horizon 2020 programme. In addition, the Commission has identified a further five key performance indicators (KPIs) giving 23 in all. The inclusion of performance indicators in the legislation is a positive step. However, we found a number of shortcomings with regard to the programme indicators, set out in the following paragraphs. Of these 23 key performance indicators, more than half measure input or output rather than results and impacts (see also Figure 3.4). Furthermore there are ambiguities in the formulation of some indicators, which make their interpretation difficult without further clarification. Examples include the Number of institutional change actions promoted by the programme or the Number of occurrences of tangible specific impacts on European policies resulting from technical and scientific support provided by the Joint Research Centre. Ambiguities in formulation may always exist, especially in legislation, so there are often further clarifications. For example, the Number of occurrences of tangible specific impacts on European policies resulting from technical and scientific support provided by the Joint Research Centre is further defined in an indicator fiche, which also includes a reference to the methodology for measuring the indicator. For a third of the Horizon 2020 key performance indicators, the Commission did not set either baselines, milestones, or targets on the grounds that these indicators represent a new approach, and therefore no targets can be set before the first results are available, due at the beginning of 2018. Only five of the 23 Horizon 2020 key performance indicators have both a baseline and a target. In our view, 2018 is much too late to be useful for a programme lasting from 2014 to 2020. Where targets do exist, they are sometimes unambitious. In some cases, targets are set below the level of results achieved under FP7. In addition, we found five Horizon 2020 indicators in the 2015 programme statements where the target was below the baseline. While the Legislative Financial Statement accompanying the Commission’s proposal for Horizon 2020 generally provided baselines and targets for indicators, a number of new indicators were included in the legislative process, and for these there was no baseline on which to consider a reasonable target. Setting a target without adequate evidence may lead to them lacking relevance and credibility (two of the RACER criteria). Horizon 2020 targets that are below the FP7 results are explained. For example, for the number of researchers who have access to research infrastructures through union support has been reduced because priority will be given to new emerging infrastructures and the targeting of new communities (starting communities) whose infrastructures are usually not able to provide as large an access as advanced communities (annual report of DG RTD, page 33). In many areas targets are ambitious and increased from FP7 — for example SME involvement, climate change related expenditure, expenditure on renewable energy, time taken to sign grants, to give some examples. For several indicators, no meaningful data will be available before 2018/2019 for reporting. For a quarter of the Horizon 2020 key performance indicators, the Commission considers that it is not meaningful to report results before 2019, because a critical mass of finished projects will not be reached before then. While these indicators might be useful for the ex post evaluation of Horizon 2020, they are not suitable for interim evaluations or for annual monitoring purposes before 2018. This is confirmed in the Horizon 2020 Monitoring Report 2014. Compared to other funding areas, there can be a considerable time lag between the moment research funds are disbursed and the moment research produces outputs and the moment when these outputs produce results (during and after the project’s lifetime). For example, at the time of the ex post evaluation of FP7 published in January 2016 (hence, 2 years after the end of FP7), more than half of FP7 projects were still ongoing. For this reason, the Commission will continue reporting on FP7 results in a dedicated section of the monitoring reports of Horizon 2020. Moreover, the interim evaluation of Horizon 2020 will include an assessment of the longer term impacts of previous framework programmes. This approach will continue over time mitigating this problem. A practical example of the problems caused by the poor use of indicators is provided in our special report on the EIT. We noted that the organisation’s core indicators continued to focus on input or output rather than results or impacts, and concluded that it was difficult to assess the performance of the EIT since it started operations in 2010. There was a lack of robust and relevant reporting and monitoring processes and few impacts were visible. As stated in the reply of the Commission and EIT to special report No 4/2016, the EIT performs comprehensive analyses to compare the Knowledge and Innovation Communities (KICs) based on several factors, not simply on the basis of the core KPIs. The implementation of KICs’ business plans is assessed both through quantitative performance indicators and through qualitative reviews done by external experts and the EIT’s staff. More widely, the EIT also conducts on-site monitoring and topical reviews of the KICs’ activities. These tools, taken together, provide a very informative picture of the differences between the KICs. There is insufficient focus on performance at the level of work programmes and calls The work programmes and associated calls are a key driver in the way that Horizon 2020 operates. They provide an opportunity for the Commission to translate the high-level general and specific objectives of Horizon 2020 into lower level objectives useful for operational performance management purposes. Based on the specific programme, the Commission prepares 2-year work programmes that develop the specific objectives and expected impact, and set out funding opportunities through calls for proposals: 15 billion euro for 2014-2015 and 14,5 billion euro for 2016-2017. The main Horizon 2020 work programme dedicates one part to each specific objective of the specific programme. It is complemented by separate work programmes for the ERC (an annual programme), the JRC (biennial programmes), and the EIT (a triennial programme, broken down annually). Since the observation of the Court the Commission has introduced a strategic planning process that went on before the 2016-2017 work programmes were finalised. This is a strategic examination of priorities and needs, carried out with Member States and other stakeholders, to ensure the link between the Horizon 2020 specific objectives and the work programmes implementing them. The legislation for Horizon 2020 was adopted shortly before the implementation of the programme started, limiting the scope for strategic planning. The Europe 2020 strategy, the Innovation Union and other flagship initiatives are referred to in the overall strategic framework for the work programme 2014-2015, but the references remain generic, unsystematic and do not clarify the specific contribution of the individual work programmes to particular strategic objectives. For 2016-2017, the strategic programming process improved and was documented via an overarching strategic planning document and scoping papers. The preparation of the Horizon 2020 legislation was accompanied by the parallel preparation of the work programmes 2014-2015. There was a clear link of each work programme part to the Horizon 2020 specific objectives. For all the work programmes we examined, the objectives are consistent with the specific objectives as defined in the specific programme. However, the objectives in the sampled work programmes do not translate the high-level objectives from the legislation into operational ones which can be monitored and reported on: three out of five work programme objectives use exactly the same wording as is in the legislation, and none fulfil the SMART criteria. The objectives are not accompanied by indicators with targets and milestones. Operational objectives and relevant indicators would facilitate the required focus on performance or monitoring and reporting on performance at the work programme level. It is accepted that the expected impact established in the work programmes can be improved, and this has already started in the 2016-2017 work programmes. However, the primary focus of the work programme is to provide the information necessary for applicants to prepare high quality proposals. Work programmes do not need to identify their contribution to Europe 2020 objectives. They should rather be linked to the specific objectives of Horizon 2020, and these objectives have been formulated to comply with Europe 2020. Performance will firstly be assessed through the indicators set out in the legislation. This is made clear in Article 5 of the Horizon 2020 legislation. The first set of relevant performance indicators for assessing progress against the general objective is set out in [the Annex to the Regulation]. This will be the primary information on which the assessment of performance will be established. Each work programme contains calls for proposals corresponding to the broad lines of activities of the specific objective (representing the scientific and technological content). The sampled calls for proposals identify objectives which are in line with those in the work programme and consistent with specific broad lines of activities, where applicable. Although overall the objectives of the calls are more specific than the ones in the work programme, they do not meet all SMART criteria. Beside objectives, each call also includes an expected impact section. However the expected impacts are rarely quantified and generally do not include indicators. It will thus be difficult for the Commission to measure the success of the calls in the absence of operational objectives and indicators with targets. The sampled work programmes do not identify their long-term contribution to the strategic objectives even though this should be explained according to guidance on work programmes. None of the examples we analysed explain how the work programme contributes towards the achievement of specific Europe 2020 objectives; in one case, the objective of the work programme links the actions supported to one of the flagship initiatives but does not specify exactly how. THE COURT’S OBSERVATIONS There is a focus on performance in proposals, evaluations and grant agreements but there are weaknesses that can lead to problems when assessing performance Under the calls, applicants submit proposals for funding. The Commission evaluates the proposals and for those it chooses to fund the Commission signs grant agreements with beneficiaries. In this section we assess the focus on performance for these stages of the management process. Evaluation of proposals Proposals are normally evaluated by experts, on behalf of the Commission, based on their excellence, impact and the quality and efficiency of their implementation. With the exception of the ERC, which uses only the excellence criterion, the beneficiaries’ proposals we examined all included sections structured around these three evaluation criteria, with the weighting for impact at least as high as the other two criteria. However, the quality of this evaluation depends on the quality of the expected impacts set out in the calls published by the Commission. For the excellence criterion, templates included in proposals indicate how objectives should be defined. Three out of the five proposal templates sampled required applicants to define SMART objectives and explain how they relate to the topic of the relevant work programme. The objectives of these three proposals are generally SMART. However, for the other two proposals, related to the ERC and Marie Curie, the sampled proposals did not include SMART objectives as there was no such requirement. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Grant agreements Grant agreements are the main way in which Horizon 2020 funds are allocated to deliver the programme objectives. The sampled grant agreements used the same objectives as were in the proposals. Grant agreements include a section on background and results. These results mean any tangible and intangible output of the action as defined in the rules of participation. There is an important provision in the grant agreements which requires beneficiaries to report to the Commission information for the aggregation of the Horizon 2020 KPIs; this level is therefore an important source of all performance information for Horizon 2020. We have pointed to problems assessing performance at the higher level (work programmes and calls) in the previous paragraphs. See Commission’s replies to paragraphs 3.36 to 3.38. Information on the indicators established in the legal base can and will be collected to allow an assessment of performance. The design of the monitoring and reporting system of Horizon 2020 limits the Commission’s ability to properly monitor and report on performance This section focuses on the monitoring and reporting system for Horizon 2020. We examined whether: Pt_a_. 4. the current setup enables the Commission to monitor and report separately the spending on research and development as opposed to innovation; Pt_b_. 4. the contribution of Horizon 2020 to Europe 2020 can be reported on; Pt_c_. 4. the Commission will be in a position to evaluate the programme in the course of its life. Txt. 9. Spending on research is not monitored separately from that on innovation Horizon 2020 brings together previously separate funding instruments for research and innovation in a common strategic framework. In line with this approach, the innovation-related parts of the Competitiveness and Innovation Framework Programme and the EIT were brought together under the umbrella of Horizon 2020. We pointed out in the past that the proposed legislation for Horizon 2020 did not provide a clear definition of innovation and that without some limits on the scope of fundable innovation activities there was a risk that a wide range of innovation projects would dilute funding. In our view this risk persists. The framework programme and the specific programme for Horizon 2020 do not provide clear criteria to distinguish between research and development and innovation or to monitor and report separately on the performance under these headings. One of the major aims of Horizon 2020, set out from the beginning of discussions on the framework programme, was to narrow the divide between research and innovation. The legislation gives no separate definition of research and innovation instead stating that: Research and innovation activities mean the whole spectrum of activities of research, technological development, demonstration and innovation, including the promotion of cooperation with third countries and international organisations, the dissemination and optimisation of results and the stimulation of high quality training and mobility of researchers in the Union. The Commission considers that distinguishing between research and innovation activities is not required, has no clear added value, and risks to run counter to the efforts to narrow the gap between research and innovation in Europe. THE COURT’S OBSERVATIONS The EU has adopted the 3 % headline target for the research and development intensity indicator. Not all of the innovation activity funded by Horizon 2020 falls within the definition of research and development for the purpose of this target. As a matter of fact, there are activities within Horizon 2020 that aim specifically at fostering innovation as opposed to research and development and, therefore, are not captured by the 3 % target. The contribution of Horizon 2020 to Europe 2020 whilst established in the budget will not be reported on in meaningful way at least until 2018 The programme statements, used in the budgetary procedure to justify operational expenditure, provide information on the links between spending programmes and the Europe 2020 strategy by signalling the headline targets pursued and by estimating the financial contribution of programmes to the flagship initiatives. Therefore, the financing of the Europe 2020 strategy from the EU budget is calculated based on the estimated financial contribution of spending programmes to the flagship initiatives in the programme statements. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES In terms of the budgetary process, through the programme statements Horizon 2020 links over 90 % of its 2015 budget to the different flagship initiatives. Almost half of this sum (4,3 billion out of 9,5 billion euro) was allocated to actions related to flagship initiative Innovation Union. While links between Horizon 2020 and Europe 2020 are established for the budgetary procedure, they are not reflected in the performance reporting on Horizon 2020. The Horizon 2020 legislation requires monitoring to include information on cross-cutting topics, one of which is the contribution to all relevant Europe 2020 flagship initiatives. However, the Horizon 2020 annual monitoring report 2014 (published in April 2016) does not address the contribution of Horizon 2020 to the relevant flagship initiatives. The links of Horizon 2020 and Europe 2020 will only be assessed in the interim evaluation report of Horizon 2020, expected in 2017. And none of the Commission’s evaluation reports required under Article 318 of the TFEU over the past years have reported on this contribution in a useful way. The links between Horizon 2020 and the Europe 2020 strategy will be assessed in depth in the interim evaluation of Horizon 2020. The interim evaluation of Horizon 2020 will also assess the leverage issue. The Commission considers that Article 31 of the Regulation establishing Horizon 2020, which establishes the scope of the annual monitoring report, does not require that it addresses the contribution of Horizon 2020 to the relevant flagship initiatives of the Europe 2020 strategy. The monitoring report is a report on the implementation of Horizon 2020 only. According to Article 32 of the Regulation establishing Horizon 2020, the contribution of Horizon 2020 to the objectives of the Europe 2020 strategy will be taken into account in the interim evaluation of Horizon 2020. The Commission’s evaluation reports under Article 318 of the TFEU can only report on the contribution of Horizon 2020 to Europe 2020 when this information is available after the interim evaluation. Some inherent difficulties encountered in the ex post evaluation of FP7 may persist also for Horizon 2020 There are inherent difficulties in terms of timing concerning evaluations of research framework programmes. The framework programme for Horizon 2020 states that evaluations should be carried out in a sufficiently timely manner to feed into the decision-making process. This requirement is difficult to fulfil in the area of research, where results and impacts might take many years to appear. Nonetheless, the legislation sets deadlines for different evaluations: an interim evaluation is required typically in the fourth year of the 7-year framework programme (2017 for Horizon 2020), and an ex post evaluation is requested 2-3 years following programme end (2023). While these deadlines are designed to support the decision-making process, the risk is that they lead to evaluations which do not adequately capture the results and impacts of programmes. The Commission has recognised the inherent difficulty of providing meaningful evaluations within a fixed 7-year timeframe for each framework programme. This is why the Horizon 2020 interim evaluation will include an updated assessment of the performance of FP7. This will mitigate the problems identified. The Commission also recognises the need to balance the need to provide timely inputs into new legislation against the needs to evaluate the longer term effects of the added-value of research spending. This is reflected in the legislation. The Commission agrees that an interim evaluation of a research framework programme will face some inherent difficulties both in terms of outputs and in terms of data availability. However, leaving it until later may delay the learning of important lessons that can certainly be assessed at the stage of the interim evaluation. This is recognised in the legislation, as many of the specific elements in Article 32(3) that the interim evaluation is required to address can be assessed at a relatively early stage of the programme, for example assessing: (27) the participation of participants that have at their disposal high-end research infrastructures or have a history of using full-costing in the seventh framework programme, (28) the simplification for participants that have at their disposal high-end research infrastructures or have a history of using full-costing in the seventh framework programme, (29) extent of use of the additional remuneration to personnel as referred to in Article 27 of Regulation (EU) No 1290/2013. Txt. 10. The Commission considers therefore that the legislation, and evaluation practices, mitigates as far as possible the inherent difficulties recognised by the Court. The ex post evaluation of FP7 (2007-2013) was finalized towards the end of 2015, when more than half of FP7 projects were still ongoing and two years after the adoption of the successor programme, Horizon 2020. This evaluation was affected by delays in many FP7 projects and because the beneficiaries’ obligation to report stopped when the projects ended. Furthermore, the data collection for some FP7 performance indicators ceased in 2013, when the seventh framework period ended. For Horizon 2020, the interim evaluation due to be undertaken in 2017 will face constraints because no meaningful results will be available for several performance indicators (see paragraph 3.31 above). And evaluation in general for Horizon 2020 is likely to be affected by the shortcomings identified in reporting results for calls and work programmes (where the data collected from projects cannot be aggregated to give feedback on calls and work programmes — see paragraphs 3.33 to 3.38); and by the difficulty the Commission faces in collecting performance information from the beneficiaries after the closure of projects. The Commission may carry out evaluations of the impact after the projects end, but in the cases we examined we found no requirements for beneficiaries to report after the closure of projects. The Commission can aggregate information to assess the 23 indicators set out in the legal base, and this is the cornerstone of the performance framework, as set out in Article 5 of the Horizon 2020 Regulation: The first set of relevant performance indicators for assessing progress against the general objective is set out in [the Annex to the legislation]. The Commission does not always use key performance concepts consistently, with the risk that the performance management system is undermined Figure 3.1 sets out the various ways in which EU action can lead to benefits: output, result and impact. Outcome includes both results and impacts. These terms are used as part of the Commission’s performance monitoring and reporting system to define objectives and indicators. Any inconsistencies or weaknesses in the way that the terms are used risks undermining a performance management system. The Commission broadly uses these terms in line with these definitions. However, in Horizon 2020, the rules for participation by beneficiaries define results as any tangible or intangible output of the action, such as data, knowledge or information that is generated in the action. This definition would normally be applied to output rather than result. The adoption of the Horizon 2020 Regulations (December 2013) preceded the better regulation guidelines (May 2015). This has led to some inconsistencies, as noted by the Court. Since the adoption of the better regulation guidance, the better regulation terminology is consistently used in the monitoring and evaluation of the performance of Horizon 2020. Another inconsistency in the application of these terms in Horizon 2020 is that, under the legislation, the individual work programmes which drive Horizon 2020 should include expected results whereas the criterion used in the rules of participation is impacts. However, the terminology used by the Commission in work programmes and calls for proposals is expected impacts. Results measure the immediate effects directly attributable to an action, and are thus much more suitable as criteria for selecting projects than impacts, which are longer term and may include elements which cannot be addressed at project level. Our sample of five ongoing Horizon 2020 projects shows that the notion of expected impact was interpreted differently under the various calls. One beneficiary, for the SME scheme, was required to submit a market analysis and a business plan. These documents were part of the proposal and analysed by the evaluators for the impact criterion. Another beneficiary provided a brief qualitative description of the link between the project and the various expected impacts, such as Increased innovation opportunities for service providers, including SMEs and public administrations. A third beneficiary described the benefits of a new treatment protocol that would become available as a result of the project, and confirmed that the results should contribute to the impact defined in the relevant call. The diverse interpretation of expected impact in calls, and the different ways beneficiaries found to justify their contribution, poses the risk of inadequately assessing project performance in relation to the overall goals of calls and work programmes. The need to strengthen the expected impact statements was identified as an issue for the second work programme and the existing guidance was improved. The quality of the evaluation process will still ensure that best projects are selected. Regarding the beneficiary that was required to submit a market analysis and a business plan, this is for the SME scheme and a market analysis and business plan are conditions for the grant. The emphasis of Horizon 2020 is on innovation and longer term impacts. This is why expected impact has been used in the calls, and is assessed during evaluations. The use of the word results in the legislation arises because the Horizon 2020 legislation preceded the better regulation guidelines. The contribution of projects in different areas is assessed using a range of indicators set out in the Horizon 2020 legal acts and other documents. And indicators are only one source of information. For the monitoring and evaluation of performance, the Commission uses a wide range of data to assess this (e.g. through external studies, surveys, case studies, counterfactual analysis etc.). Another problem stemming from the Commission’s lack of clarity in this area is that some of the indicators described result or impact indicators actually deal with input and output therefore limiting the Commission’s ability to assess performance. Some examples of how these risks have materialised are in Figure 3.4. The indicators set out in the Regulation cover input, output, results and impacts. The Commission considers that monitoring of these indicators, backed up by evaluation, will provide a good basis on which to assess performance. Inconsistent use of indicators As paragraph 3.28 notes, there are 23 Horizon 2020 key performance indicators for assessing the results and impacts of the specific objectives of Horizon 2020. Seven of these indicators are linked to publications, such as Leadership in enabling and industrial technologies (LEIT) — Number of joint public-private publications. Achieving outputs such as this may be a valid objective of the programme, but such publication indicators do not measure either results or impacts. In addition, there are four other indicators that are linked to input rather than result or impact. Examples include Access to risk finance — Total investments mobilised via debt financing and venture capital investments and the Percentage of the overall Energy challenge funds allocated to the following research activities: renewable energy, end user energy-efficiency, smart grids and energy storage activities. We also reported on this issue in chapter 3 of the last year’s annual report, in the context of the ESI funds. Inconsistent use of indicators The Commission’s better regulation guidelines have established a standard definition of these terms and these definitions will now be applied consistently. While publications in themselves are outputs, the Commission considers that in certain cases they can be result indicators for example, publications in peer reviewed journals. The Commission considers that its performance framework, and especially the indicators used for Horizon 2020, provide an appropriate balance between different types of indicators. Within this balance, a certain number of input indicators is acceptable, to ensure that different priorities are adequately assessed. PART 2 — PERFORMANCE PLANNING AND REPORTING AT DIRECTORATE-GENERAL LEVEL: THE NATURAL RESOURCES FAMILY This part of the chapter covers performance reporting at the level of Commission DGs. We reviewed the 2015 programme statements, management plans and annual activity reports of four DGs, based on requirements set out in the Financial Regulation, the Commission's internal control standards and its internal instructions relating to MPs and AARs. According to the Commission’s 2015 MP instructions, DGs should establish general objectives with impact indicators (long-term) and specific objectives with result indicators (short/medium term). According to the Financial Regulation, DGs’ objectives should be SMART. In line with our observations in previous years in relation to other DGs, we found that, for all four DGs reviewed, many of the objectives (general and specific) used in the MPs and AARs were taken directly from policy or legislative documents and were thus at too high a level to be useful as management instruments (see Figure 3.5). Since the Court’s observation in relation to the management plans for 2015, the Commission has introduced changes to the SPP cycle notably to provide a clearer framework for the Commission’s accountability and for its services. With the new approach, the planning documents as from 2016 are meant to be more streamlined and focused on the priorities of the Commission and the competencies of the DGs, helping them to establish priorities and tell a coherent performance story with a clear distinction between attribution (for results and outputs that are a direct consequence of the Commission services’ activities) and contribution (where the Commission services’ activities play a part in bringing about a given result or an impact) for EU high level policy and spending programmes’ results. The strategic plan introduces a new approach shifting the focus from short-term planning to a long-term, strategic perspective helping the DGs to align their specific objectives with the long-term objectives defined at Commission. The operational planned outputs for each directorate-general for the year ahead are listed in the Management Plan. Examples of objectives not fit for management purposes DG CLIMA: The objective To secure investment for climate related issues is not specific as it does not indicate the volume of investment to be secured, and it is not time-bound as there is no deadline to achieve it. DG ENV: To turn the EU into a resource-efficient, green and competitive economy is set as a specific objective. It does not define exactly what the terms used mean, nor does it state how the objective should be achieved. Examples of objectives not fit for management purposes DG CLIMA — The amount and deadline are implicitly defined as the objective and measuring indicators clearly relate to the (climate action sub-programme) of the LIFE programme 2014-2020. DG ENV — The Commission notes that the way of how this objective will be achieved is set out on pages 9 and 10 of the Strategic Plan for 2016-2020 of the Directorate-General for Environment and the indicators to measure progress are set out on pages 19 to 21 of the same document. In total we reviewed 19 objectives of which only four fulfilled all the SMART criteria. This result is similar to what we have found in previous years. While the objectives examined were relevant to the policy area in which the DGs operate, in many cases the indicators accompanying the objective do not exhaustively measure all aspects at the appropriate level. The choice of indicators should be based on considerations such as relevance, measurability, and the availability of timely and reliable performance data. As far as is possible, indicators should meet the RACER criteria. Indicators which measure outcomes over which DGs have limited influence should be complemented by other indicators measuring directly the DGs’ activities. All four DGs have at least one performance indicator per objective to help management monitor, evaluate and report on achievements. Of the 58 indicators in the DGs we examined, 28 fulfilled all the RACER criteria. This is an improvement compared to what we have reported in previous years. However, as in previous years, we have identified some shortcomings in the indicators being used (see Figure 3.6). Example of problems with indicators Indicators presented as result indicators but were instead input-oriented, for example in DG AGRI Total investment in renewable energy production — This indicator measures the activity directly realised by the intervention instead of focusing on the increase in renewable energy resulting from the investment. Example of problems with indicators This indicator takes into account not only the EAFRD input but also the national and private contributions to the operations. Therefore, it provides an indication of the triggering effects of the EU/national contributions on private contributions in operations directly addressing the renewable energy objective. The 2023 quantified objective is set up ex ante by the Member States based on a needs analysis of the programming area and is meant to be achieved through a consistent strategy. In addition, and above all, this indicator is complemented by the result indicator renewable energy produced from supported projects to be measured by evaluators in 2017, 2019 and ex post evaluation (2024). Jointly taken, the two indicators will provide a complete picture on results achieved. THE COURT’S OBSERVATIONS PART 3 — RESULTS OF THE COURT’S AUDIT ON PERFORMANCE 2015 special reports and other output on performance Our special reports examine whether the principles of sound financial management are applied to EU spending. Within the framework of our 2013-2017 strategy, each year we identify a number of priorities on which to focus. We choose topics for special reports based on a range of criteria, such as the level of income or spending involved, the risks to sound financial management and the degree of stakeholder interest. In 2015, we adopted 25 special reports, as listed in Annex 3.1. Our approach to performance audit naturally evolves to reflect the changing circumstances in which the EU operates. In this annual report, we draw attention to some of the ways that our 2015 special reports take account of these changes: the development of a series of reports examining different aspects of the financial crisis; a series of reports on the chronic problem of youth unemployment; and our response to developments in climate change and the environment. EU responses to the financial crisis The financial and later sovereign debt crisis which hit in 2007-2008 has had far-reaching consequences for Europe, and the Member States of the EU and the Union as a whole have taken a number of steps in response. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES In 2015, we published three special reports in this area. Our report on financial assistance to countries in difficulties was a comprehensive analysis of the Commission’s management of the financial assistance provided to five Member States: Hungary, Latvia, Romania, Ireland and Portugal. Our conclusions were wide-ranging, drawing attention to the fact that, before the crisis, some warning signs went unnoticed by the Commission, with the result that it found itself unprepared when requests for financial assistance started to arrive. The Commission’s performance in then mobilising resources to deal with the crisis was mixed, although the Commission was generally successful in issuing debt on behalf of the EU to meet financing needs. The impact of the Commission’s action was that the Member States mostly met their deficit targets and generally continued with the reforms that had been sparked by the programme conditions. We made nine recommendations, covering all aspects of this issue. The magnitude and nature of the 2007-2008 global financial crisis was unprecedented. The crisis showed that the pre-crisis surveillance framework was not fully adequate in identifying the risk in the underlying fiscal positions in times of unprecedented economic crisis. It highlighted the need for a wider approach to surveillance in the banking sector and wider macroeconomy. At the point when countries applied for assistance, behind-the-scenes monitoring of a deteriorating situation had been underway and the Commission was in close contact with the national authorities. The response to the crisis was immediate and comprehensive. Despite acute staff resource constraints and a steep learning curve, the Commission was successful in quickly taking on all its new programme management duties, not only in issuing debt to meet financing needs in all cases. This is acknowledged by the Court, which qualifies this as an achievement. All programmes achieved their core economic objectives. The audit recommendations focused on process, decision-making and record-keeping aspects of the Commission’s programme management and the Commission accepted, or partially accepted, all the recommendations. Greece was one of the Member States most associated with the financial crisis. We therefore published a report focusing specifically on the support provided by the Commission to help Greece implement financial reform. We concluded that the body set up by the Commission to deliver its support, the Task Force for Greece, had achieved its mandate overall in delivering assistance in line with the requirements of the Greek economic adjustment programmes. However, there were shortcomings and mixed results in terms of its influence on the progress of reforms. The Commission acknowledges that the impact of technical assistance on the overall progress of reforms was mixed. This needs to be seen in the broader political context and taking into account the division of competences between Member States and the European Union Institutions. The implementation of the (structural) reforms was and remains the responsibility of the Greek authorities, whereas the Task Force for Greece was available to advise/assist the Greek authorities where requested. The role of credit rating agencies assumed greater significance as a result of the crisis, leading to the establishment of the European Securities and Markets Authority (ESMA) in 2011. Our report on the supervision of credit rating authorities by the EU was therefore an important contribution to more effective management in this area. We concluded that, in a short period of time, ESMA had laid down good foundations for carrying out its supervisory role, but that there remained room for improvement in, for example, clarifying methodological criteria and the adequacy of summary statistics disclosed in and the checks carried out relating to the data reported to ESMA’s online repository. Youth unemployment Chronic youth unemployment is a major problem in a number of Member States, and was an issue we identified as one of the priorities for our work in 2015. In response, in 2013 the Council recommended that a Youth Guarantee be established by the Member States, with a combination of EU and national funding. Under the Guarantee, young people, within four months of leaving education or losing a job, should find a good quality job offer or training to help them find employment in the future. In our report on this topic, we found that the Commission had provided timely and appropriate support to Member States in setting up their guarantee schemes, but pointed to three potential risks that might affect the schemes’ success: the adequacy of total funding, how a good quality offer should be defined, and the Commission’s monitoring and reporting on results. Concerning the three potential risks that might affect the youth guarantee schemes’ success identified by the Court, the Commission agrees and would like to point out that: (30) investing in a youth guarantee is crucial for the EU to preserve its future growth potential. Significant EU financial support is already available, but to make the youth guarantee a reality, Member States also need to prioritise youth employment measures in their national budgets, (31) good-quality is a crucial element for implementing the youth guarantee. The Commission will further stimulate the reflection around good quality offers and provide further guidance to Member States, (32) a lengthy consensus building exercise was necessary to define a common monitoring framework. Monitoring activities will feed into the Commission’s assessment of the youth guarantee implementation under future European Semester and its 2016 report on the implementation of the scheme to the European Parliament and the Council through the appropriate channels. Txt. 11. We returned to this issue in our report later in 2015 on the Commission’s support for youth action teams, an initiative which brought national and Commission experts together in the eight Member States with the highest levels of youth unemployment. We concluded that the effectiveness of the advisory support provided by the Commission to the Member States in relation to the youth action teams was limited by the level and quality of information available to it on how existing ESF or national measures for young unemployed people work on the ground. Moreover, the Commission’s assessment of the amendments to their programmes proposed by the Member States focused mainly on budgetary aspects rather than on the best use of the ESF funds available. The Commission proactively used all the instruments available to it, given the legal framework in place, to achieve the objectives of the YATs initiative and to report on them. In the Commission’s view, the YATs initiative has met the political expectations set out by the European Council in its conclusions of 30 January 2012. The key objective of the joint YATs — an ad hoc, punctual initiative — was to mobilise EU and national instruments, including structural funds, and to discuss policy reforms to be included in youth job plans. The teams were meant to shake the tree by raising awareness at the highest political level, creating a sense of urgency and giving a stronger steer while not creating additional administrative or/and legal procedures. The environment, energy and climate change The environment and climate change, together with its implications for EU policies in a number of other areas, such as agriculture and energy, is one of the global challenges we are targeting in our 2013-2017 strategy. So our report on EU-funded waste water treatment plants in the Danube river basin examined whether EU support was helping Member States achieve its waste water policy objectives — a key element of a successful environmental policy. We concluded that while EU financial support had played an important role in bringing forward water collection and treatment, this had not been sufficient to meet the relevant deadlines. A later report in 2015 then examined a different aspect of water quality in the Danube basin: progress by the same four Member States in meeting the requirements of the water framework directive, whose main objective from 2000 was to achieve good water quality by 2015 (with some exceptions possible). We concluded that the implementation of measures co-funded by the EU had led to little improvement in water quality. Indeed, while the introduction of river basin management plans as a tool for integrated water management had been a strength of the directive, in practice the Member States’ plans suffered from shortcomings in the identification of pollution pressures and in the definition of remedial measures. On a different theme, we examined the EU’s emissions trading scheme (ETS). This is an essential pillar of the EU’s policy to combat climate change, and is designed to promote reductions of greenhouse gas emissions in a cost-effective and economically efficient way. We found that management of the scheme by the Commission and the Member States was not adequate in all respects. It was hindered by some issues related to the robustness of the framework for protecting its integrity, such as the absence of EU level oversight of the emissions market, and by significant weaknesses in the implementation of the second phase of the scheme in the period 2008-2012, after the initial preparatory phase. The Commission is committed to continue improving the EU ETS; it will address the shortcomings identified by the auditors. A number of recommendations are to be implemented by Member States. Many points noted by the ECA were specific to the period 2008-12 (phase 2) and have already largely been addressed in the current rules for the period 2013-2020 (phase 3). Since the audit report, there has been a new legislative proposal on ETS (implementation of the climate and energy objectives for 2030 following the October 2014 European Council conclusions). We also published a report about the security of energy supply and the development of the EU internal energy market. We found that the EU’s objective of completing the internal energy market by 2014 had not been reached. Energy infrastructure in Europe is generally not well designed for fully integrated markets and therefore does not currently provide effective security of energy supply. Financial support from the EU budget in the field of energy infrastructure has made only a limited contribution to the internal energy market and security of energy supply. The Commission is taking concrete steps to remove the remaining market barriers to the internal energy market, in particular through regional cooperation frameworks involving different Member States and in the context of preparation of its market design initiative. With regard to infrastructure, it is important to note that: Pt_a_. 5. energy infrastructure is expected to be financed by the market, i.e. user tariffs approved by independent regulators; financial support from EU budget therefore should be the exception rather than the rule; Pt_b_. 5. where financial support has been allocated, security of supply has often been significantly improved. A good example are the Baltic States where support from both EEPR and CEF to key projects has not only increased security of supply in electricity but also helped wholesale prices to converge. Txt. 12. The follow-up of special report recommendations Following up our performance audit reports is an essential element in the cycle of accountability and helps to encourage the implementation of audit recommendations by the Commission and the Member States. In our 2016 special report on the 2014 follow up exercise, we made a number of recommendations, suggesting that the Commission should bring its follow-up practices into line with relevant internal control standards. The Commission accepted these recommendations and agreed to take corrective actions. The Commission is committed to ensuring that recommendations made by the Court in its special reports and accepted by the Commission are systematically implemented and followed-up. The Commission considers that its control arrangements are in line with international control standards and best practices. However, the Commission is committed to address further challenges concerning the discharge follow-up system and procedures. THE COURT’S OBSERVATIONS We assessed the Commission’s follow up of a sample of 90 audit recommendations from 11 special reports published in the period 2011-2012. The sample was selected on the basis of a range of criteria. Of the 83 recommendations that could be verified, we concluded that the Commission fully implemented 63 % of our recommendations, 26 % were implemented in most respects, 10 % in some respects while 1 % were not implemented. Although the Commission is our principal auditee, we frequently identify matters which can best be addressed by the Member State authorities or in cooperation with the Commission, given that most EU expenditure is under shared-management arrangements with Member States. In our special reports we sometimes also address recommendations to Member States in general. We can also direct recommendations to Member States individually although this is rare. Recommendations are communicated to Member State level although formal follow-up is mixed When we make recommendations applicable to Member States in general, there are five ways in which these messages may be transmitted (see Figure 3.7). First, each report is sent to the Supreme Audit Institution (the principal national public audit body). The reports are also sent to the Council, for its consideration in the context of the discharge procedure and, on occasion, through presentations to Council working parties. Our annual activity report (which contains details of our special reports) is also sent to the parliament of each Member State, and individual Members of the Court may bring particular reports to the special attention of a national parliament or its relevant committee, in the course of their interaction with the parliament concerned. Our recommendations are also disseminated to Member States through the medium of Commission guidelines and training interventions. Communication of findings to Member States Source: European Court of Auditors. THE COURT’S OBSERVATIONS However, to date we have had relatively little information on the impact of these communication efforts in the Member States. To remedy this, we conducted a survey of paying agencies, ministries and SAIs at Member State national level on the follow up of recommendations. Four Commission DGs also completed a questionnaire. This confirmed that communication of recommendations does indeed get through to Member State level with the survey providing evidence of receipt of our messages through one or more of the above channels. In terms of impact, 80 % of respondents acknowledged at least some awareness of our reports at Member State level. However, only 13 % responded that there was a co-ordinated follow-up process for all of our recommendations in their Member State. As regards follow-up action taken, 38 % suggested that action only takes place for particular topics, while 49 % indicated that no formal follow-up action takes place. Nevertheless, some Member States provided concrete examples of embedding our recommendations in national policy and practice, mainly in the field of natural resources (see Figure 3.8). Examples of awareness and communication at Member State level Our report on The cost-effectiveness of EU Rural Development support for non-productive investments in agriculture recommended that Member States should put in place procedures to ensure that the costs of the supported non-productive investments do not exceed the costs of similar types of goods, service or works offered by the market. More concretely, Member States should define appropriate benchmarks and/or reference costs against which the costs of non-productive investments are systematically verified as part of their administrative checks. Responding to this recommendation, one Member State has introduced standard prices for the management of planting of hedges. For disseminating the Court's recommendations, a managing authority in one Member State makes the Court's findings available on its rural development programme implementation committee intranet where all the implementing bodies and the paying agency are represented. In one Member State, the Ministry of Agriculture notifies the paying agency and, where necessary, other stakeholders by means of a summary report with our key findings and recommendations, with the full report in an annex. Source: Select survey responses from our survey of SAIs and paying agencies. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The survey exercise shows that there is a reasonable level of awareness of our performance audit recommendations at Member State level. However, there is a wide variation in the level of formal follow-up and only moderate evidence of changes in national policy and practice as a result of our recommendations. It is clear that there is scope for improvement in how we develop and distribute our recommendations for maximum impact at Member State level. In the autumn of 2016, the Court will engage the Contact Committee, Commission services and the Council in the consultation on possible follow-up of recommendations in Member States. CONCLUSIONS AND RECOMMENDATIONS Conclusions Horizon 2020 is an ambitious, wide-ranging programme, bringing previously separate research and innovation programmes together for the first time, and incorporating a focus on Societal challenges. There are high-level links between Horizon 2020 and the EU’s main strategy, Europe 2020: two of the three main indicators for Horizon 2020 are also high-level indicators for Europe 2020. These indicators, however, are of limited use in tracking the contribution of Horizon 2020 to Europe 2020 (see paragraphs 3.7 to 3.17). While the Commission accepts that it needs to improve further, it considers that its performance framework already tracks the contribution of Horizon 2020 to Europe 2020. As explained under Figure 3.1 and in paragraph 3.13: The EU budget represents only a small part of total research and innovation spending in Europe, and so in itself can make only a minor contribution towards the achievement of the Europe 2020 objectives. Progress towards the Europe 2020 strategy depends on a wide range of actions at EU and national level, which are not only on spending, and for which Member States have a primordial role to play. The contribution of Horizon 2020 to the achievement of the Europe 2020 strategy is reinforced by the interaction with other actions at EU and national level. The links between the Commission’s 10 political priorities and Europe 2020/Horizon 2020 need further clarification. The political priorities overlap with, but are different from, the Europe 2020 priorities. It is not clear to which set of priorities Horizon 2020 — in its planning right through to its monitoring and reporting — is now to be aligned. There is a risk that other areas of EU activity are concerned (see paragraphs 3.18 to 3.21). The political priorities of the Juncker Commission provide the framework for the Commission’s own contribution to the delivery of the Europe 2020 strategy, which is a shared responsibility of the Union and the Member States. The political priorities guide the Commission’s implementation of financial programmes but this is done in full complementarity with the Europe 2020 strategy and in accordance with the legal framework underpinning Horizon 2020, which has not changed. A key success factor for Horizon 2020 is effective synergy and complementarity between national and European research and innovation programmes. Our partners in the supreme audit institutions of Bulgaria and Portugal found that while there were areas of complementarity in their countries, there were also some problems at the national level (see paragraphs 3.22 and 3.23). The legal framework of Horizon 2020 introduces several important elements for performance management such as objectives and key performance indicators. This is a welcome development. However, further developments to the framework are still needed; some baselines, milestones and targets still need to be developed, some targets should be reviewed to ensure that they are suitably ambitious; some indicators suffer from lack of available information. The objectives defined by the legislation are themselves naturally at a high-level, but are not translated, through work programmes and calls, into lower level objectives useful for operational performance management purposes. The Commission does not use the Horizon 2020 work programmes and associated calls to increase the focus on performance (see paragraphs 3.27 to 3.38). The performance framework was discussed with the legislative authority during the adoption of the Horizon 2020 legislation, and is set out in that legislation. It includes a number of indicators that will be monitored throughout the programme, and interim and ex post evaluations. This clear framework is an advance on previous framework programmes. The results and impacts of work programmes, and of Horizon 2020 more generally, will be assessed using this performance framework. The long-term nature of research, the timescale of the preparation of the legislation, and the limitations in the availability of some data do provide challenges in the completion of the performance framework, and the Commission accepts that continual improvements are necessary. For the proposals and grant agreements we examined, objectives were generally SMART when it was required by the Commission. The evaluation process of these proposals, with the exception of ERC that uses excellence as a criterion, had an emphasis on performance. Grant agreements require beneficiaries to report to the Commission information for the aggregation of the Horizon 2020 KPIs. However the use of the wider concept of expected impact rather than expected result increases the risk that information provided for this part is too broad and not directly attributable to the action funded by Horizon 2020, with the result that performance assessment could be difficult (see paragraphs 3.39 to 3.43). The use of expected impact rather than result arises from the fact that the Horizon 2020 legislation predates the better regulation guidelines. The Commission wants to evaluate the expected impact of projects when considering which projects to support. This will not affect the monitoring of the indicators that have been set out in the legislation, and which are the cornerstone of the performance framework. Weaknesses in the design of the monitoring and reporting system for Horizon 2020 mean that the Commission is limited in its ability to monitor and report on the performance of the programme: The Commission considers that the Horizon 2020 monitoring and reporting system is a great improvement compared to FP7. Clear objectives have been established, and a performance framework, discussed fully with stakeholders and the legislative authority, has been anchored in the legal basis. The Commission is aware of some weaknesses that remain, and is working to remedy them: Pt_a_. 6. The current setup does not enable the Commission to monitor and report separately the spending and the performance of the different components of research and development and innovation within Horizon 2020 (see paragraphs 3.45 to 3.47). Pt_a_. 7. the Commission does not consider that research and innovation spending needs to be reported on separately, this is not required by the legislation and runs counter to the efforts made in Horizon 2020 to close the gap between research and innovation; Pt_b_. 6. The contribution of Horizon 2020 to Europe 2020 is well established in the budgetary process. However, thus far the contribution of Horizon 2020 to Europe 2020 has not yet been reported on in a meaningful way (see paragraphs 3.48 to 3.50). Pt_b_. 7. the legislation requires the contribution of Horizon 2020 to Europe 2020 to be reported on in the interim evaluation of the programme, and this evaluation will be undertaken in 2017. The nature of the research cycle means that a certain time must elapse before the contribution to Europe 2020 can be reported on in a meaningful way; Pt_c_. 5. The Horizon 2020 monitoring and reporting system has improved compared to FP7 but there remain weaknesses regarding the information gathered for Horizon 2020. There is a risk that difficulties encountered in the ex post evaluation of FP 7 might persist for Horizon 2020 (see paragraphs 3.51 to 3.53). Pt_c_. 6. the Commission agrees that the Horizon 2020 monitoring and reporting system has improved compared to FP7. Clear objectives have been established, and a performance framework, discussed fully with stakeholders and the legislative authority, has been anchored in the legal basis. Txt. 13. It considers that it has acted to mitigate the difficulties identified in the FP7 evaluation system. In assessing performance, some basic concepts are the different types of benefits which can result from public action: inputs, outputs, results and impacts. While the Commission has defined how these terms should be used, it does not always abide by these definitions. A particular problem is the apparently arbitrary use of result and impact in translating the legislation into individual work programmes (see paragraphs 3.54 to 3.59). The adoption of the Horizon 2020 Regulations (December 2013) preceded the better regulation guidelines (May 2015). This has led to some inconsistencies, as noted by the Court. Since the adoption of the better regulation guidance, the better regulation terminology is consistently used in the monitoring and evaluation of the performance of Horizon 2020. For the natural resources family of DGs, we noted that many of the objectives used in MPs and AARs were taken directly from policy or legislative documents and were therefore at too high a level for management purposes. We observed progress for the indicators used to measure performance but some shortcomings remain (see paragraphs 3.60 to 3.65). Since the Court’s observation in relation to the management plans for 2015, the Commission has introduced changes to the SPP cycle notably to provide a clearer framework for the Commission’s accountability and for its services. With the new approach, the planning documents as from 2016 are meant to be more streamlined and focused on the priorities of the Commission and the competencies of the DGs, helping them to establish priorities and tell a coherent performance story with a clear distinction between attribution (for results and outputs that are a direct consequence of the Commission services’ activities) and contribution (where the Commission services’ activities play a part in bringing about a given result or an impact) for EU high level policy and spending programmes’ results. The strategic plan introduces a new approach shifting the focus from short-term planning to a long-term, strategic perspective helping the DGs to align their specific objectives with the long-term objectives defined at Commission. The operational planned outputs for the year ahead are listed in the Management Plan. Recommendations Annex 3.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, we presented five recommendations. Of these recommendations, two were implemented in some respects and three were not implemented. Our recommendations for 2015 are as follows: (41) Recommendation 1: Operational objectives The Commission should translate the Horizon 2020 legislation high-level objectives into operational objectives at work programme level, so that, by assessing the performance of work programmes and calls, they can effectively be used as drivers for performance. Txt. 14. The Commission accepts this recommendation. While the overall performance framework is laid down and clear, the Commission notes the points raised by the Court, which will be fully taken into account as part of work underway to improve the work programmes, including its role in performance based reporting. (42) Recommendation 2: Further clarification for the different strategies are needed The Commission should further clarify the links between the Europe 2020 strategy (2010-2020), the multi-annual financial framework (2014-2020), the Commission priorities (2015-2019) through, for instance, the strategic planning and reporting process (2016-2020). This would strengthen monitoring and reporting arrangements and enable the Commission to report effectively on the contribution of the EU budget towards Europe 2020 objectives. Txt. 15. The Commission accepts this recommendation. The political priorities of the Commission are clearly defined in President Juncker’s political guidelines. They provide a roadmap for the Commission’s action that is fully consistent and compatible with Europe 2020 as Europe’s long-term growth strategy. The differences in scope between the priorities and Europe 2020 reflect the Commission’s duty to respond to challenges that have emerged or grown in prominence since the development of the Europe 2020 strategy, such as migration. From 2016 onwards and based on the strategic plans produced by each Commission department, the annual activity reports of each Commission department as well as the annual management and performance report of the Commission will explain how specific objectives contribute to the political priorities and will further clarify the links with Europe 2020. (43) Recommendation 3: Consistent terminology use The Commission should, across all its activities, use the terms input, output, result and impact consistently and in line with its better regulation guidelines. Txt. 16. The Commission accepts this recommendation. The adoption of the different Regulations for the 2014-2020 MFF period preceded the better regulation guidelines (May 2015). Moreover, the content of those regulations is the result of a political negotiation between the co-legislators. This has led to some inconsistencies, as noted by the Court. Since the adoption of the better regulation guidance, the better regulation terminology should be consistently used in the monitoring and evaluation of all programmes. Consistent use of terms across all legislation may, however, not be possible in the short term. With regard to the Commission’s internal performance framework, the Commission has provided consistent definitions and methodologies through the standing instructions provided to all departments. These definitions are equivalent to those contained in the better regulation guidelines.

THE COURT’S OBSERVATIONS INTRODUCTION The subject of this chapter is performance (see Figure 3.1). It continues the theme — which has been a feature of our annual reports since 2010 — of including a chapter assessing different aspects of the systems put in place by the Commission to ensure that EU funds are spent well, and not just in line with relevant legislation. Since 2010, we have analysed the Commission’s self-assessment on performance as stated in the annual activity reports (AARs) of a selection of the Commission’s directors-general, and included an examination of some of the themes developed in that year’s special reports on performance audit. Other elements of performance that we have examined are, the management plans of selected directorates-general (DGs) and the reporting by the Commission to the European Parliament and the Council including the first few editions of the Commission’s evaluation reports. Performance in the EU Performance in the EU is assessed on the basis of the sound financial management principles (economy, efficiency and effectiveness), and covers:

(a) inputs — financial, human, material, organisational or regulatory means needed for the implementation of the programme;

(b) outputs — the deliverables of the programme;

(c) results — the immediate effects of the programme on direct addressees or recipients;

(d) impacts — long-term changes in society that are, at least partly, attributable to the EU’s action.

In the equivalent chapter of last year’s annual report, we took a particular look at the EU’s long-term strategy, Europe 2020. We examined the EU’s high-level monitoring and reporting of the strategy and, for the European Structural and Investment Funds (the ESI funds), how the strategy feeds through into the partnership agreements and funding programmes agreed between the Commission and Member States. This year we continue that theme, and focus our examination of the Commission’s performance through Horizon 2020, the Commission’s research and innovation programme and, as will be seen below, a very significant item in the EU budget. The chapter therefore consists of three parts. Horizon 2020 performance is the subject of the first part. The remaining two parts deal with subjects we have examined in previous years: the extent to which performance is covered in the main annual planning and reporting documents prepared by the Commission’s directors-general and some of the main themes arising from our 2015 special reports including the follow-up of a sample of recommendations from previous special reports and, for the first time, responses from the Member States to a survey that we conducted on their follow up of our recommendations. In the 2014 annual report, we reported that, based on information supplied by the Commission, the mid-term review of Europe 2020, which should have been available to the Council in March 2015, had been delayed until early 2016. As of June 2016, the Commission informed us that the mid-term review had been finalised and that there would now be no formal mid-term review concluding document. As a result, we are not in a position to report in this chapter on how this review process was carried out. We may, however, return to this issue in future reports. PART 1 — HORIZON 2020 (2014-2020) There are three elements to our analysis of Horizon 2020. First we set out the main features of the programme and examine the strategic framework in which it operates, in particular its links to other EU strategies. The second element considers the advantages and potential risks involved in bringing research and innovation activity together. Finally we analyse the extent to which there is a focus on performance within Horizon 2020. Section 1 — Links between Europe 2020, the new Commission priorities and Horizon 2020 do not always facilitate a focus on performance and need to be clarified Europe 2020 and Horizon 2020 Europe 2020 is the European Union’s 10-year jobs and growth strategy. It was launched in 2010 to create the conditions for smart, sustainable and inclusive growth. Responsibility for delivering the Europe 2020 strategy is shared between the EU and its 28 Member States. Thus the strategy’s five headline targets have an associated national target in each Member State, reflecting the different circumstances of each. Horizon 2020 has a budget of around 75 billion euro for 2014-2020. It is the EU’s programme for supporting research and innovation, with the general objective to contribute to building a society and an economy based on knowledge and innovation across the Union. It supports the Europe 2020 strategy and the EU objective of creating a European Research Area (ERA). It brings together previously separate spending programmes for research and innovation. Funding is provided mainly in form of grants, although access to other forms of finance, such as financial instruments, is also available. The Commission manages the majority of expenditure under Horizon 2020, either itself or through agencies and other bodies. Beneficiaries of Horizon 2020 are individuals, teams, universities, research institutes, SMEs and larger companies, from within and outside the EU. Many funded projects are collaborative. Figure 3.2 shows the structure of Horizon 2020. The general objective is translated into three priorities (1. Excellent science, 2. Industrial leadership and 3. Societal challenges). These priorities are pursued through 14 specific objectives, complemented by broad lines of activities which provide their scientific and technological content. The general objective is also pursued through two specific objectives (Spreading excellence and widening participation and Science with and for society), while two additional organisations, the JRC and the EIT also contribute to the general objective and the priorities by following their own specific objectives. Horizon 2020 also includes additional research measures, such as Fast-Track to Innovation to promote close-to-the-market innovation activities. The structure of Horizon 2020 Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES There are high-level links between Horizon 2020 and Europe 2020 but need to be enhanced to ensure that they facilitate a greater focus on performance Under the legislation, Horizon 2020 has three performance indicators for assessing progress against its general objective:

(a) the research and development intensity indicator according to which investment should account for at least 3 % of GDP by 2020 — replicating one of Europe 2020’s five headline targets;

(b) an innovation output indicator;

(c) the share of researchers in the active population. This indicator has no direct link with Europe 2020 headline targets.

The 3 % indicator measures the input of funds in research and development. It was not designed to measure output, results or impact. Horizon 2020 adds for the first time in legislation the requirement that private — as opposed to public — investment in research and development should account for two-thirds of the total. Against these targets, in 2014 total such investment across Europe was just over 2 % of GDP, with the private sector accounting for around 64 % of this. Comparable figures for the USA were 2,8 % in 2012, 3,5 % for Japan and 4,2 % for South Korea (for both countries in 2013). Whilst the 3 % R&D intensity indicator measures the input of funds in research and development, it also reflects the effectiveness of policy measures taken to stimulate a high quality science base in the Member States and to create better framework conditions for research and innovation in the private sector. Horizon 2020 spending is one relatively small contribution to the target. It was always expected that Member States would provide the bulk of funds in this area. In 2014, the first year of implementation, Horizon 2020 spending by the EU institutions represented just under 10 % of total public budgets for R&D. We noted in last year’s annual report that meeting all national targets for research and development would lead to a total investment of 2,6 % of GDP by 2020, some way below the 3 % target. In the area of research, as well as Horizon 2020, efforts are made to coordinate and improve Member States’ investment in research and innovation, and to develop framework conditions that allow private sector innovation to thrive (open innovation). This is crucial to making significant progress in closing the innovation divide between the European Union and its key competitors. Increasing impact and excellent science through openness (open science) and a digital single market, embedding digital into the grand societal challenges, implementing the research, innovation and competitiveness dimension of the Energy Union, together with a forward-looking climate-change policy, and translating Europe’s strengths in science and technology into a leading global voice (open to the world) are also key objectives of research policy. Research and innovation investments from the cohesion policy programmes for 2007-2013 also contributed to the progress towards the 3 % target. This contribution is continued for the 2014-2020 programming period thanks to the thematic concentration of ERDF and will strengthen its impact and leverage thanks to the Smart Specialisation ex ante conditionality for research and innovation. The contribution of Horizon 2020 to the achievement of the Europe 2020 strategy is reinforced by the interaction with other actions at EU and national level. The second Horizon 2020 performance indicator, on innovation, is a composite output indicator. It was initially developed for the Europe 2020 flagship initiative Innovation Union. It captures information on four elements: patents, employment in knowledge-intensive activities, competitiveness of knowledge-intensive goods and services, and employment in fast-growing firms of innovative sectors. The Composite Indicators Research Group (financed by the EU) has pointed to the fact that such indicators, based on sub-indicators with no common meaningful unit of measurement and with no obvious method for weighting, have pros and cons. In particular, while they illustrate the big picture, there is a risk that in isolation they can lead to simplistic policy conclusions; they should therefore be used in combination with the sub-indicators. The Commission reported on this indicator and the sub-indicators in March 2016. The innovation output indicator has the advantage that it focuses attention on establishing the right conditions for innovation and that it provides a measure for assessing progress in innovation at the level of the EU economy. However, it also has a number of shortcomings which limit its usefulness for monitoring the Horizon 2020 contribution towards the Europe 2020 strategy objectives: 3.15.

the Commission has not yet set a quantified target to be reached by a defined date;

In order for a quantitative target to impact policymaking and stimulate progress, the target set must be consensual, justified on the basis of evidence and benchmarking, and easy to understand and communicate. This is more difficult to achieve in the case of composite indicators, such as the Innovation Output Indicator, as the summary scores in which they result are by nature abstract. Therefore, composite indicators typically do not come with target levels. Another example of this is the Summary Index of the Innovation Union Scoreboard, which also does not have a target level, but still clearly has an impact on policymaking of Member States. Furthermore, the available time series for some of the component indicators are still too short to establish trend lines and identify a meaningful EU target level (the component on high-growth enterprise was based on a new Eurostat data collection). In addition, setting an EU target level would require benchmarking against the performance non-EU countries, something not possible at the time the indicator was launched.

the Commission has not translated it into national targets to reflect the individual situation of each Member State — as recommended by experts in the field for indicators measuring innovation.

The Research intensity target is set by Member State, and the Innovation Output Indicator complements it by providing the policy picture on innovation output. These different elements need to be seen together.

The specific objectives set by Horizon 2020 are the next level in the hierarchy of objectives where links are established to the Europe 2020 strategy. The largest element (38 %) of the Horizon 2020 budget is allocated to the priority Societal challenges which is designed to address policy priorities in the Europe 2020 strategy. In many cases, the description of the specific objectives refers to the flagship initiatives they are designed to support. However, although they are a useful indication of the generic relationship between Horizon 2020 and the Europe 2020 strategy, they are not further translated at lower levels into more specific and operational links that could serve as the basis for monitoring and reporting the contribution of Horizon 2020 to the individual flagship initiatives and thus the Europe 2020 strategy. For the preparation of the budget, the contribution of Horizon 2020 to each flagship initiative is provided in the Horizon 2020 programme statement on an annual basis, following a methodology developed with the launch of the programme. The links between the Commission’s 10 political priorities and Europe 2020/Horizon 2020 need further clarification In 2013, the regulation establishing Horizon 2020 set priorities, general and specific objectives for the period 2014-2020. These priorities and objectives were designed to be aligned with, and to contribute to, the Europe 2020 strategy. Towards the end of 2014, the current Commission set 10 priorities for the Commission as a whole for 2014-2019. While these priorities highlight some new areas (such as a free trade agreement with the United States), they also overlap significantly, as one would expect, with the existing Europe 2020 strategy, including its headline targets and flagship initiatives. For example, employment and climate change are issues which feature heavily in both. Despite the overlaps and complementarities, it has been difficult to establish the extent to which the 10 political priorities form a distinct new approach of the Commission or are intended as a new way for the Commission to put Europe 2020 into operation for its work. The Europe 2020 strategy is the growth strategy for a decade for the EU and at EU level. In line with the Europe 2020 strategy, the 10 political priorities of the Commission provide a clear and focused agenda/strategy for the activities of the Commission towards Europe 2020 and they are entirely consistent with the overall Europe 2020 goals. At the same time, they provide greater clarity and focus in the Commission’s overall political strategy, taking into account recent developments. For example, since the establishment of the Europe 2020 strategy the EU has seen sovereign debt crises, increasing public sector deficits, war in Ukraine and Syria, a massive flow of migrants towards it, and the COP 21 climate agreement. These 10 political priorities have also been presented to the European Parliament, the European Council and the Council in the course of the appointment of the new Commission. The priorities take over, as targets and indicators, the targets and indicators of the Europe 2020 strategy. It is normal for strategic priorities to alter, as circumstances change. However, as pointed out by the Commission’s internal audit service, although there is common ground between the two strategies, the Commission has, so far, not mapped the correspondence between them. In our view, since the relationship between the two strategies has not been clarified, they coexist and are used for different purposes: The Commission considers that its priorities provide a clear and focused policy agenda for the Commission that is consistent with the Europe 2020 strategy for the European Union as a whole. As such, they are naturally used for different purposes. However, this does not mean that they are in any way inconsistent or incoherent, they are entirely complementary. Please refer also to the Commission reply to paragraph 3.18.

The allocation of the EU budget to spending programmes (such as Horizon 2020) is based on, among other things, their contribution to the Europe 2020 headline targets and flagship initiatives. The monitoring and evaluation of spending programmes also builds on objectives designed to reflect the Europe 2020 strategy.

In relation to Horizon 2020 this was implemented in the first work programmes, for 2014-2015. Over the remainder of the programme the work programmes, and the budget for particular elements of the work programme, will be increasingly influenced by the political priorities set out by the current Commission in response to the emerging political, economic and technological context.

The Commission’s annual work programme now focuses on the 10 political priorities. Starting from 2016, the Commission has also introduced a new method of strategic planning, which requires DGs to prepare 5-year strategic plans, also based on the 10 political priorities. The 5-year strategic plans will become the starting point for drafting annual management plans (MPs) for carrying out the DGs’ tasks (including the implementation of spending programmes such as Horizon 2020). Since the reporting in annual activity reports is strongly linked to the MPs, annual activity reports from 2016 onwards can also be expected to reflect primarily the 10 political priorities.

The Commission’s annual work programme includes the actions that the Commission will take to help deliver on the political priorities and contribute to the achievement of the Europe 2020 strategy. The new multiannual Strategic Plans developed for each Commission department are built around the 10 political priorities, expressed as general objectives. As part of this exercise, the DGs were instructed to explain how their activities contribute to the Europe 2020 strategy. To facilitate this, the headline targets of the Europe 2020 strategy appear as impact indicators linked to the relevant general objectives. The accompanying annual management plans set out the planned outputs for each department and how these will contribute to the objectives. Commission departments will report on their achievements through their annual activity reports, which will reflect the structure of the strategic and management plans. However, the Commission will continue to report on progress towards the Europe 2020 targets as part of the European Semester process and on the contribution of the budget to Europe 2020 in the annual management and performance report.

For Horizon 2020, this leads to a situation where the legal framework and the budget allocation for the programme reflect the Europe 2020 strategy, while those parts of the Commission implementing Horizon 2020 have refocused strategic planning and management arrangements on the 10 political priorities. For example, the key priorities to be developed in the Horizon 2020 work programmes 2016-2017 are based on seven out of the 10 policy areas on the agenda of the new Commission. Under these circumstances, it is unclear what strategic objectives Horizon 2020 should be targeting. DG RTD identified four political priorities where the DG can make a contribution. The pursuit of these priorities, however, will require a refocusing of effort against the original targeting as set out in the Horizon 2020 legislation. The Commission considers that there is no lack of clarity about which strategic objectives Horizon 2020 should be targeting. These objectives are the Europe 2020 objectives, as set out in the legal base. The 10 political priorities are used to consider how the Commission can best assist in reaching the Europe 2020 priorities. For example, the delivery of the digital single market (DSM) strategy will lead to updated priorities in the work programme 2017-2018 of the ICT part of Horizon 2020 in line with both Europe 2020 and the digital single market Commission priority, within the flexibility available in Horizon 2020. The respective set of measures and initiatives of 18 April 2016 (DSM package) already demonstrates this alignment of Horizon 2020 and the DSM priority. In last year’s annual report we pointed to the weak alignment between the Europe 2020 and the EU’s seven year budgetary cycle. The same complication may affect the 10 political priorities, as the mandate of the current Commission ends in 2019, a year before the end date of Europe 2020 and Horizon 2020. The 2007-2013 MFF period may not have fully coincided with the period covered by the Europe 2020 strategy. However, the Commission maintains its position formulated in reply to paragraphs 3.7-3.11, 3.87 of the Court’s annual report for 2014, that this does not imply a lack of alignment with the overall strategy. Moreover, the 2014-2020 MFF period falls entirely within, and has the same end date, as the current Europe 2020 strategy period. The 2014-2020 MFF has been designed with a view to concentrate on delivering on the Europe 2020 strategy. The Commission is monitoring and reporting on the contribution of the EU budget to the EU’s overall objectives and to the Europe 2020 strategy in particular. The current Multiannual Financial Framework, Horizon 2020 and the Europe 2020 strategy are fully aligned and have a common end date of 2020. These frameworks are not affected by the expiry of the mandate of the current Commission in 2019. Section 2 — Synergies and complementarities with ESI funds and national programmes Limited information on synergies with ESI funds One of the main objectives of the Europe 2020 strategy and its Innovation Union is the completion of the ERA in which researchers, scientific knowledge and technology circulate freely. The EU budget contributes to the ERA objective through Horizon 2020 and the ESI funds. In this context, and in line with the legislation, the Commission envisaged a system for encouraging and tracking synergies between Horizon 2020 and the ESI funds. The classification of Horizon 2020 projects according to ESI thematic objectives, together with the introduction of specific references to potential synergies at their planning and reporting stages would provide the basis for such a system. However, in the first Horizon 2020 monitoring report, information on the use and impact of synergies between Horizon 2020 and the ESI funds is limited. The Commission plans to analyse the impact and synergies between Horizon 2020 and the ESIF in the context of the Horizon 2020 interim evaluation. The Commission has created the regulatory conditions that can enable synergies between the funds by increased coordination and harmonisation of the relevant legal frameworks. However, the implementation of synergies needs a continued effort both on the side of the bodies in charge of Horizon 2020 and of the authorities of the Member States and regions. Significant efforts were undertaken in Member States and regions, in particular in the framework of the development of smart specialisation strategies to develop policy measures that allow to implement the different synergy types identified by the Commission’s guidance and to focus R&I efforts on fields where the Member State or region has the biggest chances for excellence in terms of science and business competitiveness. These fields also offer the biggest chance to be successful in Horizon 2020 calls. Also on the side of Horizon 2020 management, extensive action is being undertaken to promote and develop synergies, for example through the set-up of memoranda of understanding between joint undertakings or the seal of excellence initiative, launched at the initiative by Commissioners Moedas and Creţu in October 2015. A number of regions and Member States are already involved in this initiative. The Commission also implements the Stairway to Excellence initiative helping the EU-13 Member States to improve their access to Horizon 2020. Examples of complementarities between national and EU research programmes The Horizon 2020 legislation recognises the importance of building synergies and complementarities between national and European research and innovation programmes. Figure 3.3 below includes key messages from our cooperation with the Supreme Audit Institutions (SAIs) of Bulgaria and Portugal: while there are some areas in which national and EU research programmes are complementary, our partners also found some problems at the national level. Examples of complementarities with national programmes including some identified weaknesses at national level Examples of complementarities with national programmes including some identified weaknesses at national level The relation between the EU priorities in the context of the Europe 2020 strategy for smart, sustainable and inclusive growth, including Horizon 2020, and the national priorities of Bulgaria, is articulated through eight priorities set in the National Development Programme Bulgaria 2020 (NDP). Three of these priorities are related to science and innovation:

Improving the access to and enhancing the quality of education and training and the quality characteristics of the workforce.

Support for innovation and investment activities to increase the competitiveness of the economy.

Energy security and increased resource efficiency, particularly for SMEs.

In addition to the NDP, which is an integrated document for the socio-economic development of Bulgaria up to 2020, the objectives under Horizon 2020 are also synchronised and broken down into operational objectives and plans in:

the National Strategy of Scientific Research (NSSR) to 2020, which is designed to support scientific development in Bulgaria and make it into a factor for the development of an economy based on knowledge and innovation activities,

the Innovation Strategy for Smart Specialization (ISSS), that has the strategic objective for Bulgaria to move from the group of modest innovators to the group of moderate innovators by 2020.

The national strategies set a national objective to increase the share of expenditure on research and development activity to 1,5 % of GDP by 2020. To achieve the national objective measures and activities are envisaged in the Action Plans implementing the national strategic documents. However, the BNAO identified that neither quantitative indicators for results, nor target values per year for activities related to Horizon 2020 are included in the action plans for the NDP. Neither were quantitative indicators per activity included in the action plan for the NSSR. The BNAO has also identified some issues regarding coordination and interaction between all the participants engaged in Horizon 2020 at the national level, and that improving the interaction between all the participants would stimulate Bulgaria’s participation in the programme. The Commission supports Bulgaria actively both via the Stairway to Excellence pilot (see above) and the new lagging regions initiative in order to improve its innovation eco-system and participation in Horizon 2020 via the implementation and evolution of the Bulgarian smart specialisation strategy. The European Commission presented in October 2015 key policy recommendations to help modernise the research and innovation sectors in Bulgaria. This was the first time that an EU Member State made voluntary use of the Horizon 2020 Policy Support Facility (PSF), a new instrument that gives governments practical support to identify, implement and evaluate those reforms needed to enhance the quality of their public research and innovation structures. The main national strategic document in the area of research and innovation is the Research and Innovation Strategy for Smart Specialization (RISSS) that has a 2014-2020 timeframe, matching the Horizon 2020 timeframe and Europe 2020 end date. One of its chapters includes links between Europe 2020 and the Portugal 2020 partnership agreement, signed between Portugal and the Commission, in the area of R&D and innovation, mobilizing funds from both cohesion policies and Horizon 2020. The Portuguese Tribunal de Contas noted that the RISSS included two scenarios for the evolution of the indicator on the proportion of GDP invested in research and development: a conservative one of 2 % achieved by 2020, and a change one with 2,7 % by 2020. It also found that, taking account of Europe 2020, Horizon 2020 indicators and the Innovation Union indicators, some issues are not tackled by the national strategy indicators, such as the researchers’ gender, the contribution that is sustainable development and climate change related, and the participation of SMEs. The Tribunal de Contas also noted that there are similar indicators at national and European level, such as those dealing with scientific publications or innovative products or processes, that cannot be compared, for instance because different denominators are used. Section 3 — The mechanics of Horizon 2020 do not consistently drive a focus on performance Measuring performance in the area of research is inherently difficult. The results and impacts of research by their nature are difficult to predict, generally rather long-term and often do not follow a linear path. However, the EU still needs to obtain reliable information on scientific progress, productivity and quality for the research it finances, in order to measure scientific performance for good decision making. This need is recognised in the Horizon 2020 framework regulation which states that effective performance management, including evaluation and monitoring, requires the development of specific performance indicators […] which are relevant to the appropriate hierarchy of objectives and activities. Against this background, this section examines whether:

(a) there is a focus on performance, including a clear and operational link to Europe 2020, in the grant management process of Horizon 2020;

(b) the design of the monitoring and reporting system of Horizon 2020 enables the Commission to monitor and report on performance; and

(c) important basic concepts of performance monitoring and reporting are used in a consistent and correct manner by the Commission.

The grant management processes are not fully designed to focus on performance To assess whether the Horizon 2020 grant management processes included an appropriate focus on performance, particularly in its links to Europe 2020, we examined the design of the Horizon 2020 legal framework, the focus on performance at the level of work programme and calls and, for each of the five work programmes we sampled, one successful proposal by beneficiaries. The legal framework of Horizon 2020 introduces several important elements for performance management but there are some weaknesses in design, particularly as regards the use of performance indicators The objectives and indicators agreed by the legislative authority for Horizon 2020 represent a real improvement on the previous framework programmes. As we have observed in the past, objectives defined in legislation may be at too general and high a level for operational use by management. This is also the case for these high-level Horizon 2020 objectives — even those referred to as specific objectives in the regulation, most of which are not actually formulated as objectives but rather as challenges and/or areas. For example, for the priority Societal challenges two specific objectives are: health, demographic change and well-being; and secure clean and efficient energy. The extent to which these high-level objectives are successfully translated into lower-level objectives by management is examined in the next section. For Horizon 2020, the inclusion of objectives in the legal text, for the first time, was an undoubted step forward. It did lead to an extensive debate with the legislative authority about the objectives of the programme, and the indicators to be monitored. As well as specifying the objectives of the programme, the Horizon 2020 legislation also includes three performance indicators for assessing progress against the general objective, set out in paragraph 3.10. At the next level down of the legal framework, the specific programme introduces further elements important for performance management. It lists 18 key indicators for assessing results and impacts of the specific objectives of the Horizon 2020 programme. In addition, the Commission has identified a further five key performance indicators (KPIs) giving 23 in all. The inclusion of performance indicators in the legislation is a positive step. However, we found a number of shortcomings with regard to the programme indicators, set out in the following paragraphs. Of these 23 key performance indicators, more than half measure input or output rather than results and impacts (see also Figure 3.4). Furthermore there are ambiguities in the formulation of some indicators, which make their interpretation difficult without further clarification. Examples include the Number of institutional change actions promoted by the programme or the Number of occurrences of tangible specific impacts on European policies resulting from technical and scientific support provided by the Joint Research Centre. Ambiguities in formulation may always exist, especially in legislation, so there are often further clarifications. For example, the Number of occurrences of tangible specific impacts on European policies resulting from technical and scientific support provided by the Joint Research Centre is further defined in an indicator fiche, which also includes a reference to the methodology for measuring the indicator. For a third of the Horizon 2020 key performance indicators, the Commission did not set either baselines, milestones, or targets on the grounds that these indicators represent a new approach, and therefore no targets can be set before the first results are available, due at the beginning of 2018. Only five of the 23 Horizon 2020 key performance indicators have both a baseline and a target. In our view, 2018 is much too late to be useful for a programme lasting from 2014 to 2020. Where targets do exist, they are sometimes unambitious. In some cases, targets are set below the level of results achieved under FP7. In addition, we found five Horizon 2020 indicators in the 2015 programme statements where the target was below the baseline. While the Legislative Financial Statement accompanying the Commission’s proposal for Horizon 2020 generally provided baselines and targets for indicators, a number of new indicators were included in the legislative process, and for these there was no baseline on which to consider a reasonable target. Setting a target without adequate evidence may lead to them lacking relevance and credibility (two of the RACER criteria). Horizon 2020 targets that are below the FP7 results are explained. For example, for the number of researchers who have access to research infrastructures through union support has been reduced because priority will be given to new emerging infrastructures and the targeting of new communities (starting communities) whose infrastructures are usually not able to provide as large an access as advanced communities (annual report of DG RTD, page 33). In many areas targets are ambitious and increased from FP7 — for example SME involvement, climate change related expenditure, expenditure on renewable energy, time taken to sign grants, to give some examples. For several indicators, no meaningful data will be available before 2018/2019 for reporting. For a quarter of the Horizon 2020 key performance indicators, the Commission considers that it is not meaningful to report results before 2019, because a critical mass of finished projects will not be reached before then. While these indicators might be useful for the ex post evaluation of Horizon 2020, they are not suitable for interim evaluations or for annual monitoring purposes before 2018. This is confirmed in the Horizon 2020 Monitoring Report 2014. Compared to other funding areas, there can be a considerable time lag between the moment research funds are disbursed and the moment research produces outputs and the moment when these outputs produce results (during and after the project’s lifetime). For example, at the time of the ex post evaluation of FP7 published in January 2016 (hence, 2 years after the end of FP7), more than half of FP7 projects were still ongoing. For this reason, the Commission will continue reporting on FP7 results in a dedicated section of the monitoring reports of Horizon 2020. Moreover, the interim evaluation of Horizon 2020 will include an assessment of the longer term impacts of previous framework programmes. This approach will continue over time mitigating this problem. A practical example of the problems caused by the poor use of indicators is provided in our special report on the EIT. We noted that the organisation’s core indicators continued to focus on input or output rather than results or impacts, and concluded that it was difficult to assess the performance of the EIT since it started operations in 2010. There was a lack of robust and relevant reporting and monitoring processes and few impacts were visible. As stated in the reply of the Commission and EIT to special report No 4/2016, the EIT performs comprehensive analyses to compare the Knowledge and Innovation Communities (KICs) based on several factors, not simply on the basis of the core KPIs. The implementation of KICs’ business plans is assessed both through quantitative performance indicators and through qualitative reviews done by external experts and the EIT’s staff. More widely, the EIT also conducts on-site monitoring and topical reviews of the KICs’ activities. These tools, taken together, provide a very informative picture of the differences between the KICs. There is insufficient focus on performance at the level of work programmes and calls The work programmes and associated calls are a key driver in the way that Horizon 2020 operates. They provide an opportunity for the Commission to translate the high-level general and specific objectives of Horizon 2020 into lower level objectives useful for operational performance management purposes. Based on the specific programme, the Commission prepares 2-year work programmes that develop the specific objectives and expected impact, and set out funding opportunities through calls for proposals: 15 billion euro for 2014-2015 and 14,5 billion euro for 2016-2017. The main Horizon 2020 work programme dedicates one part to each specific objective of the specific programme. It is complemented by separate work programmes for the ERC (an annual programme), the JRC (biennial programmes), and the EIT (a triennial programme, broken down annually). Since the observation of the Court the Commission has introduced a strategic planning process that went on before the 2016-2017 work programmes were finalised. This is a strategic examination of priorities and needs, carried out with Member States and other stakeholders, to ensure the link between the Horizon 2020 specific objectives and the work programmes implementing them. The legislation for Horizon 2020 was adopted shortly before the implementation of the programme started, limiting the scope for strategic planning. The Europe 2020 strategy, the Innovation Union and other flagship initiatives are referred to in the overall strategic framework for the work programme 2014-2015, but the references remain generic, unsystematic and do not clarify the specific contribution of the individual work programmes to particular strategic objectives. For 2016-2017, the strategic programming process improved and was documented via an overarching strategic planning document and scoping papers. The preparation of the Horizon 2020 legislation was accompanied by the parallel preparation of the work programmes 2014-2015. There was a clear link of each work programme part to the Horizon 2020 specific objectives. For all the work programmes we examined, the objectives are consistent with the specific objectives as defined in the specific programme. However, the objectives in the sampled work programmes do not translate the high-level objectives from the legislation into operational ones which can be monitored and reported on: three out of five work programme objectives use exactly the same wording as is in the legislation, and none fulfil the SMART criteria. The objectives are not accompanied by indicators with targets and milestones. Operational objectives and relevant indicators would facilitate the required focus on performance or monitoring and reporting on performance at the work programme level. It is accepted that the expected impact established in the work programmes can be improved, and this has already started in the 2016-2017 work programmes. However, the primary focus of the work programme is to provide the information necessary for applicants to prepare high quality proposals. Work programmes do not need to identify their contribution to Europe 2020 objectives. They should rather be linked to the specific objectives of Horizon 2020, and these objectives have been formulated to comply with Europe 2020. Performance will firstly be assessed through the indicators set out in the legislation. This is made clear in Article 5 of the Horizon 2020 legislation. The first set of relevant performance indicators for assessing progress against the general objective is set out in [the Annex to the Regulation]. This will be the primary information on which the assessment of performance will be established. Each work programme contains calls for proposals corresponding to the broad lines of activities of the specific objective (representing the scientific and technological content). The sampled calls for proposals identify objectives which are in line with those in the work programme and consistent with specific broad lines of activities, where applicable. Although overall the objectives of the calls are more specific than the ones in the work programme, they do not meet all SMART criteria. Beside objectives, each call also includes an expected impact section. However the expected impacts are rarely quantified and generally do not include indicators. It will thus be difficult for the Commission to measure the success of the calls in the absence of operational objectives and indicators with targets. The sampled work programmes do not identify their long-term contribution to the strategic objectives even though this should be explained according to guidance on work programmes. None of the examples we analysed explain how the work programme contributes towards the achievement of specific Europe 2020 objectives; in one case, the objective of the work programme links the actions supported to one of the flagship initiatives but does not specify exactly how. THE COURT’S OBSERVATIONS There is a focus on performance in proposals, evaluations and grant agreements but there are weaknesses that can lead to problems when assessing performance Under the calls, applicants submit proposals for funding. The Commission evaluates the proposals and for those it chooses to fund the Commission signs grant agreements with beneficiaries. In this section we assess the focus on performance for these stages of the management process. Evaluation of proposals Proposals are normally evaluated by experts, on behalf of the Commission, based on their excellence, impact and the quality and efficiency of their implementation. With the exception of the ERC, which uses only the excellence criterion, the beneficiaries’ proposals we examined all included sections structured around these three evaluation criteria, with the weighting for impact at least as high as the other two criteria. However, the quality of this evaluation depends on the quality of the expected impacts set out in the calls published by the Commission. For the excellence criterion, templates included in proposals indicate how objectives should be defined. Three out of the five proposal templates sampled required applicants to define SMART objectives and explain how they relate to the topic of the relevant work programme. The objectives of these three proposals are generally SMART. However, for the other two proposals, related to the ERC and Marie Curie, the sampled proposals did not include SMART objectives as there was no such requirement. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Grant agreements Grant agreements are the main way in which Horizon 2020 funds are allocated to deliver the programme objectives. The sampled grant agreements used the same objectives as were in the proposals. Grant agreements include a section on background and results. These results mean any tangible and intangible output of the action as defined in the rules of participation. There is an important provision in the grant agreements which requires beneficiaries to report to the Commission information for the aggregation of the Horizon 2020 KPIs; this level is therefore an important source of all performance information for Horizon 2020. We have pointed to problems assessing performance at the higher level (work programmes and calls) in the previous paragraphs. See Commission’s replies to paragraphs 3.36 to 3.38. Information on the indicators established in the legal base can and will be collected to allow an assessment of performance. The design of the monitoring and reporting system of Horizon 2020 limits the Commission’s ability to properly monitor and report on performance This section focuses on the monitoring and reporting system for Horizon 2020. We examined whether:

(a) the current setup enables the Commission to monitor and report separately the spending on research and development as opposed to innovation;

(b) the contribution of Horizon 2020 to Europe 2020 can be reported on;

(c) the Commission will be in a position to evaluate the programme in the course of its life.

Spending on research is not monitored separately from that on innovation Horizon 2020 brings together previously separate funding instruments for research and innovation in a common strategic framework. In line with this approach, the innovation-related parts of the Competitiveness and Innovation Framework Programme and the EIT were brought together under the umbrella of Horizon 2020. We pointed out in the past that the proposed legislation for Horizon 2020 did not provide a clear definition of innovation and that without some limits on the scope of fundable innovation activities there was a risk that a wide range of innovation projects would dilute funding. In our view this risk persists. The framework programme and the specific programme for Horizon 2020 do not provide clear criteria to distinguish between research and development and innovation or to monitor and report separately on the performance under these headings. One of the major aims of Horizon 2020, set out from the beginning of discussions on the framework programme, was to narrow the divide between research and innovation. The legislation gives no separate definition of research and innovation instead stating that: Research and innovation activities mean the whole spectrum of activities of research, technological development, demonstration and innovation, including the promotion of cooperation with third countries and international organisations, the dissemination and optimisation of results and the stimulation of high quality training and mobility of researchers in the Union. The Commission considers that distinguishing between research and innovation activities is not required, has no clear added value, and risks to run counter to the efforts to narrow the gap between research and innovation in Europe. THE COURT’S OBSERVATIONS The EU has adopted the 3 % headline target for the research and development intensity indicator. Not all of the innovation activity funded by Horizon 2020 falls within the definition of research and development for the purpose of this target. As a matter of fact, there are activities within Horizon 2020 that aim specifically at fostering innovation as opposed to research and development and, therefore, are not captured by the 3 % target. The contribution of Horizon 2020 to Europe 2020 whilst established in the budget will not be reported on in meaningful way at least until 2018 The programme statements, used in the budgetary procedure to justify operational expenditure, provide information on the links between spending programmes and the Europe 2020 strategy by signalling the headline targets pursued and by estimating the financial contribution of programmes to the flagship initiatives. Therefore, the financing of the Europe 2020 strategy from the EU budget is calculated based on the estimated financial contribution of spending programmes to the flagship initiatives in the programme statements. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES In terms of the budgetary process, through the programme statements Horizon 2020 links over 90 % of its 2015 budget to the different flagship initiatives. Almost half of this sum (4,3 billion out of 9,5 billion euro) was allocated to actions related to flagship initiative Innovation Union. While links between Horizon 2020 and Europe 2020 are established for the budgetary procedure, they are not reflected in the performance reporting on Horizon 2020. The Horizon 2020 legislation requires monitoring to include information on cross-cutting topics, one of which is the contribution to all relevant Europe 2020 flagship initiatives. However, the Horizon 2020 annual monitoring report 2014 (published in April 2016) does not address the contribution of Horizon 2020 to the relevant flagship initiatives. The links of Horizon 2020 and Europe 2020 will only be assessed in the interim evaluation report of Horizon 2020, expected in 2017. And none of the Commission’s evaluation reports required under Article 318 of the TFEU over the past years have reported on this contribution in a useful way. The links between Horizon 2020 and the Europe 2020 strategy will be assessed in depth in the interim evaluation of Horizon 2020. The interim evaluation of Horizon 2020 will also assess the leverage issue. The Commission considers that Article 31 of the Regulation establishing Horizon 2020, which establishes the scope of the annual monitoring report, does not require that it addresses the contribution of Horizon 2020 to the relevant flagship initiatives of the Europe 2020 strategy. The monitoring report is a report on the implementation of Horizon 2020 only. According to Article 32 of the Regulation establishing Horizon 2020, the contribution of Horizon 2020 to the objectives of the Europe 2020 strategy will be taken into account in the interim evaluation of Horizon 2020. The Commission’s evaluation reports under Article 318 of the TFEU can only report on the contribution of Horizon 2020 to Europe 2020 when this information is available after the interim evaluation. Some inherent difficulties encountered in the ex post evaluation of FP7 may persist also for Horizon 2020 There are inherent difficulties in terms of timing concerning evaluations of research framework programmes. The framework programme for Horizon 2020 states that evaluations should be carried out in a sufficiently timely manner to feed into the decision-making process. This requirement is difficult to fulfil in the area of research, where results and impacts might take many years to appear. Nonetheless, the legislation sets deadlines for different evaluations: an interim evaluation is required typically in the fourth year of the 7-year framework programme (2017 for Horizon 2020), and an ex post evaluation is requested 2-3 years following programme end (2023). While these deadlines are designed to support the decision-making process, the risk is that they lead to evaluations which do not adequately capture the results and impacts of programmes. The Commission has recognised the inherent difficulty of providing meaningful evaluations within a fixed 7-year timeframe for each framework programme. This is why the Horizon 2020 interim evaluation will include an updated assessment of the performance of FP7. This will mitigate the problems identified. The Commission also recognises the need to balance the need to provide timely inputs into new legislation against the needs to evaluate the longer term effects of the added-value of research spending. This is reflected in the legislation. The Commission agrees that an interim evaluation of a research framework programme will face some inherent difficulties both in terms of outputs and in terms of data availability. However, leaving it until later may delay the learning of important lessons that can certainly be assessed at the stage of the interim evaluation. This is recognised in the legislation, as many of the specific elements in Article 32(3) that the interim evaluation is required to address can be assessed at a relatively early stage of the programme, for example assessing:

the participation of participants that have at their disposal high-end research infrastructures or have a history of using full-costing in the seventh framework programme,

the simplification for participants that have at their disposal high-end research infrastructures or have a history of using full-costing in the seventh framework programme,

extent of use of the additional remuneration to personnel as referred to in Article 27 of Regulation (EU) No 1290/2013.

The Commission considers therefore that the legislation, and evaluation practices, mitigates as far as possible the inherent difficulties recognised by the Court. The ex post evaluation of FP7 (2007-2013) was finalized towards the end of 2015, when more than half of FP7 projects were still ongoing and two years after the adoption of the successor programme, Horizon 2020. This evaluation was affected by delays in many FP7 projects and because the beneficiaries’ obligation to report stopped when the projects ended. Furthermore, the data collection for some FP7 performance indicators ceased in 2013, when the seventh framework period ended. For Horizon 2020, the interim evaluation due to be undertaken in 2017 will face constraints because no meaningful results will be available for several performance indicators (see paragraph 3.31 above). And evaluation in general for Horizon 2020 is likely to be affected by the shortcomings identified in reporting results for calls and work programmes (where the data collected from projects cannot be aggregated to give feedback on calls and work programmes — see paragraphs 3.33 to 3.38); and by the difficulty the Commission faces in collecting performance information from the beneficiaries after the closure of projects. The Commission may carry out evaluations of the impact after the projects end, but in the cases we examined we found no requirements for beneficiaries to report after the closure of projects. The Commission can aggregate information to assess the 23 indicators set out in the legal base, and this is the cornerstone of the performance framework, as set out in Article 5 of the Horizon 2020 Regulation: The first set of relevant performance indicators for assessing progress against the general objective is set out in [the Annex to the legislation]. The Commission does not always use key performance concepts consistently, with the risk that the performance management system is undermined Figure 3.1 sets out the various ways in which EU action can lead to benefits: output, result and impact. Outcome includes both results and impacts. These terms are used as part of the Commission’s performance monitoring and reporting system to define objectives and indicators. Any inconsistencies or weaknesses in the way that the terms are used risks undermining a performance management system. The Commission broadly uses these terms in line with these definitions. However, in Horizon 2020, the rules for participation by beneficiaries define results as any tangible or intangible output of the action, such as data, knowledge or information that is generated in the action. This definition would normally be applied to output rather than result. The adoption of the Horizon 2020 Regulations (December 2013) preceded the better regulation guidelines (May 2015). This has led to some inconsistencies, as noted by the Court. Since the adoption of the better regulation guidance, the better regulation terminology is consistently used in the monitoring and evaluation of the performance of Horizon 2020. Another inconsistency in the application of these terms in Horizon 2020 is that, under the legislation, the individual work programmes which drive Horizon 2020 should include expected results whereas the criterion used in the rules of participation is impacts. However, the terminology used by the Commission in work programmes and calls for proposals is expected impacts. Results measure the immediate effects directly attributable to an action, and are thus much more suitable as criteria for selecting projects than impacts, which are longer term and may include elements which cannot be addressed at project level. Our sample of five ongoing Horizon 2020 projects shows that the notion of expected impact was interpreted differently under the various calls. One beneficiary, for the SME scheme, was required to submit a market analysis and a business plan. These documents were part of the proposal and analysed by the evaluators for the impact criterion. Another beneficiary provided a brief qualitative description of the link between the project and the various expected impacts, such as Increased innovation opportunities for service providers, including SMEs and public administrations. A third beneficiary described the benefits of a new treatment protocol that would become available as a result of the project, and confirmed that the results should contribute to the impact defined in the relevant call. The diverse interpretation of expected impact in calls, and the different ways beneficiaries found to justify their contribution, poses the risk of inadequately assessing project performance in relation to the overall goals of calls and work programmes. The need to strengthen the expected impact statements was identified as an issue for the second work programme and the existing guidance was improved. The quality of the evaluation process will still ensure that best projects are selected. Regarding the beneficiary that was required to submit a market analysis and a business plan, this is for the SME scheme and a market analysis and business plan are conditions for the grant. The emphasis of Horizon 2020 is on innovation and longer term impacts. This is why expected impact has been used in the calls, and is assessed during evaluations. The use of the word results in the legislation arises because the Horizon 2020 legislation preceded the better regulation guidelines. The contribution of projects in different areas is assessed using a range of indicators set out in the Horizon 2020 legal acts and other documents. And indicators are only one source of information. For the monitoring and evaluation of performance, the Commission uses a wide range of data to assess this (e.g. through external studies, surveys, case studies, counterfactual analysis etc.). Another problem stemming from the Commission’s lack of clarity in this area is that some of the indicators described result or impact indicators actually deal with input and output therefore limiting the Commission’s ability to assess performance. Some examples of how these risks have materialised are in Figure 3.4. The indicators set out in the Regulation cover input, output, results and impacts. The Commission considers that monitoring of these indicators, backed up by evaluation, will provide a good basis on which to assess performance. Inconsistent use of indicators As paragraph 3.28 notes, there are 23 Horizon 2020 key performance indicators for assessing the results and impacts of the specific objectives of Horizon 2020. Seven of these indicators are linked to publications, such as Leadership in enabling and industrial technologies (LEIT) — Number of joint public-private publications. Achieving outputs such as this may be a valid objective of the programme, but such publication indicators do not measure either results or impacts. In addition, there are four other indicators that are linked to input rather than result or impact. Examples include Access to risk finance — Total investments mobilised via debt financing and venture capital investments and the Percentage of the overall Energy challenge funds allocated to the following research activities: renewable energy, end user energy-efficiency, smart grids and energy storage activities. We also reported on this issue in chapter 3 of the last year’s annual report, in the context of the ESI funds. Inconsistent use of indicators The Commission’s better regulation guidelines have established a standard definition of these terms and these definitions will now be applied consistently. While publications in themselves are outputs, the Commission considers that in certain cases they can be result indicators for example, publications in peer reviewed journals. The Commission considers that its performance framework, and especially the indicators used for Horizon 2020, provide an appropriate balance between different types of indicators. Within this balance, a certain number of input indicators is acceptable, to ensure that different priorities are adequately assessed. PART 2 — PERFORMANCE PLANNING AND REPORTING AT DIRECTORATE-GENERAL LEVEL: THE NATURAL RESOURCES FAMILY This part of the chapter covers performance reporting at the level of Commission DGs. We reviewed the 2015 programme statements, management plans and annual activity reports of four DGs, based on requirements set out in the Financial Regulation, the Commission's internal control standards and its internal instructions relating to MPs and AARs. According to the Commission’s 2015 MP instructions, DGs should establish general objectives with impact indicators (long-term) and specific objectives with result indicators (short/medium term). According to the Financial Regulation, DGs’ objectives should be SMART. In line with our observations in previous years in relation to other DGs, we found that, for all four DGs reviewed, many of the objectives (general and specific) used in the MPs and AARs were taken directly from policy or legislative documents and were thus at too high a level to be useful as management instruments (see Figure 3.5). Since the Court’s observation in relation to the management plans for 2015, the Commission has introduced changes to the SPP cycle notably to provide a clearer framework for the Commission’s accountability and for its services. With the new approach, the planning documents as from 2016 are meant to be more streamlined and focused on the priorities of the Commission and the competencies of the DGs, helping them to establish priorities and tell a coherent performance story with a clear distinction between attribution (for results and outputs that are a direct consequence of the Commission services’ activities) and contribution (where the Commission services’ activities play a part in bringing about a given result or an impact) for EU high level policy and spending programmes’ results. The strategic plan introduces a new approach shifting the focus from short-term planning to a long-term, strategic perspective helping the DGs to align their specific objectives with the long-term objectives defined at Commission. The operational planned outputs for each directorate-general for the year ahead are listed in the Management Plan. Examples of objectives not fit for management purposes DG CLIMA: The objective To secure investment for climate related issues is not specific as it does not indicate the volume of investment to be secured, and it is not time-bound as there is no deadline to achieve it. DG ENV: To turn the EU into a resource-efficient, green and competitive economy is set as a specific objective. It does not define exactly what the terms used mean, nor does it state how the objective should be achieved. Examples of objectives not fit for management purposes DG CLIMA — The amount and deadline are implicitly defined as the objective and measuring indicators clearly relate to the (climate action sub-programme) of the LIFE programme 2014-2020. DG ENV — The Commission notes that the way of how this objective will be achieved is set out on pages 9 and 10 of the Strategic Plan for 2016-2020 of the Directorate-General for Environment and the indicators to measure progress are set out on pages 19 to 21 of the same document. In total we reviewed 19 objectives of which only four fulfilled all the SMART criteria. This result is similar to what we have found in previous years. While the objectives examined were relevant to the policy area in which the DGs operate, in many cases the indicators accompanying the objective do not exhaustively measure all aspects at the appropriate level. The choice of indicators should be based on considerations such as relevance, measurability, and the availability of timely and reliable performance data. As far as is possible, indicators should meet the RACER criteria. Indicators which measure outcomes over which DGs have limited influence should be complemented by other indicators measuring directly the DGs’ activities. All four DGs have at least one performance indicator per objective to help management monitor, evaluate and report on achievements. Of the 58 indicators in the DGs we examined, 28 fulfilled all the RACER criteria. This is an improvement compared to what we have reported in previous years. However, as in previous years, we have identified some shortcomings in the indicators being used (see Figure 3.6). Example of problems with indicators Indicators presented as result indicators but were instead input-oriented, for example in DG AGRI Total investment in renewable energy production — This indicator measures the activity directly realised by the intervention instead of focusing on the increase in renewable energy resulting from the investment. Example of problems with indicators This indicator takes into account not only the EAFRD input but also the national and private contributions to the operations. Therefore, it provides an indication of the triggering effects of the EU/national contributions on private contributions in operations directly addressing the renewable energy objective. The 2023 quantified objective is set up ex ante by the Member States based on a needs analysis of the programming area and is meant to be achieved through a consistent strategy. In addition, and above all, this indicator is complemented by the result indicator renewable energy produced from supported projects to be measured by evaluators in 2017, 2019 and ex post evaluation (2024). Jointly taken, the two indicators will provide a complete picture on results achieved. THE COURT’S OBSERVATIONS PART 3 — RESULTS OF THE COURT’S AUDIT ON PERFORMANCE 2015 special reports and other output on performance Our special reports examine whether the principles of sound financial management are applied to EU spending. Within the framework of our 2013-2017 strategy, each year we identify a number of priorities on which to focus. We choose topics for special reports based on a range of criteria, such as the level of income or spending involved, the risks to sound financial management and the degree of stakeholder interest. In 2015, we adopted 25 special reports, as listed in Annex 3.1. Our approach to performance audit naturally evolves to reflect the changing circumstances in which the EU operates. In this annual report, we draw attention to some of the ways that our 2015 special reports take account of these changes: the development of a series of reports examining different aspects of the financial crisis; a series of reports on the chronic problem of youth unemployment; and our response to developments in climate change and the environment. EU responses to the financial crisis The financial and later sovereign debt crisis which hit in 2007-2008 has had far-reaching consequences for Europe, and the Member States of the EU and the Union as a whole have taken a number of steps in response. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES In 2015, we published three special reports in this area. Our report on financial assistance to countries in difficulties was a comprehensive analysis of the Commission’s management of the financial assistance provided to five Member States: Hungary, Latvia, Romania, Ireland and Portugal. Our conclusions were wide-ranging, drawing attention to the fact that, before the crisis, some warning signs went unnoticed by the Commission, with the result that it found itself unprepared when requests for financial assistance started to arrive. The Commission’s performance in then mobilising resources to deal with the crisis was mixed, although the Commission was generally successful in issuing debt on behalf of the EU to meet financing needs. The impact of the Commission’s action was that the Member States mostly met their deficit targets and generally continued with the reforms that had been sparked by the programme conditions. We made nine recommendations, covering all aspects of this issue. The magnitude and nature of the 2007-2008 global financial crisis was unprecedented. The crisis showed that the pre-crisis surveillance framework was not fully adequate in identifying the risk in the underlying fiscal positions in times of unprecedented economic crisis. It highlighted the need for a wider approach to surveillance in the banking sector and wider macroeconomy. At the point when countries applied for assistance, behind-the-scenes monitoring of a deteriorating situation had been underway and the Commission was in close contact with the national authorities. The response to the crisis was immediate and comprehensive. Despite acute staff resource constraints and a steep learning curve, the Commission was successful in quickly taking on all its new programme management duties, not only in issuing debt to meet financing needs in all cases. This is acknowledged by the Court, which qualifies this as an achievement. All programmes achieved their core economic objectives. The audit recommendations focused on process, decision-making and record-keeping aspects of the Commission’s programme management and the Commission accepted, or partially accepted, all the recommendations. Greece was one of the Member States most associated with the financial crisis. We therefore published a report focusing specifically on the support provided by the Commission to help Greece implement financial reform. We concluded that the body set up by the Commission to deliver its support, the Task Force for Greece, had achieved its mandate overall in delivering assistance in line with the requirements of the Greek economic adjustment programmes. However, there were shortcomings and mixed results in terms of its influence on the progress of reforms. The Commission acknowledges that the impact of technical assistance on the overall progress of reforms was mixed. This needs to be seen in the broader political context and taking into account the division of competences between Member States and the European Union Institutions. The implementation of the (structural) reforms was and remains the responsibility of the Greek authorities, whereas the Task Force for Greece was available to advise/assist the Greek authorities where requested. The role of credit rating agencies assumed greater significance as a result of the crisis, leading to the establishment of the European Securities and Markets Authority (ESMA) in 2011. Our report on the supervision of credit rating authorities by the EU was therefore an important contribution to more effective management in this area. We concluded that, in a short period of time, ESMA had laid down good foundations for carrying out its supervisory role, but that there remained room for improvement in, for example, clarifying methodological criteria and the adequacy of summary statistics disclosed in and the checks carried out relating to the data reported to ESMA’s online repository. Youth unemployment Chronic youth unemployment is a major problem in a number of Member States, and was an issue we identified as one of the priorities for our work in 2015. In response, in 2013 the Council recommended that a Youth Guarantee be established by the Member States, with a combination of EU and national funding. Under the Guarantee, young people, within four months of leaving education or losing a job, should find a good quality job offer or training to help them find employment in the future. In our report on this topic, we found that the Commission had provided timely and appropriate support to Member States in setting up their guarantee schemes, but pointed to three potential risks that might affect the schemes’ success: the adequacy of total funding, how a good quality offer should be defined, and the Commission’s monitoring and reporting on results. Concerning the three potential risks that might affect the youth guarantee schemes’ success identified by the Court, the Commission agrees and would like to point out that:

investing in a youth guarantee is crucial for the EU to preserve its future growth potential. Significant EU financial support is already available, but to make the youth guarantee a reality, Member States also need to prioritise youth employment measures in their national budgets,

good-quality is a crucial element for implementing the youth guarantee. The Commission will further stimulate the reflection around good quality offers and provide further guidance to Member States,

a lengthy consensus building exercise was necessary to define a common monitoring framework. Monitoring activities will feed into the Commission’s assessment of the youth guarantee implementation under future European Semester and its 2016 report on the implementation of the scheme to the European Parliament and the Council through the appropriate channels.

We returned to this issue in our report later in 2015 on the Commission’s support for youth action teams, an initiative which brought national and Commission experts together in the eight Member States with the highest levels of youth unemployment. We concluded that the effectiveness of the advisory support provided by the Commission to the Member States in relation to the youth action teams was limited by the level and quality of information available to it on how existing ESF or national measures for young unemployed people work on the ground. Moreover, the Commission’s assessment of the amendments to their programmes proposed by the Member States focused mainly on budgetary aspects rather than on the best use of the ESF funds available. The Commission proactively used all the instruments available to it, given the legal framework in place, to achieve the objectives of the YATs initiative and to report on them. In the Commission’s view, the YATs initiative has met the political expectations set out by the European Council in its conclusions of 30 January 2012. The key objective of the joint YATs — an ad hoc, punctual initiative — was to mobilise EU and national instruments, including structural funds, and to discuss policy reforms to be included in youth job plans. The teams were meant to shake the tree by raising awareness at the highest political level, creating a sense of urgency and giving a stronger steer while not creating additional administrative or/and legal procedures. The environment, energy and climate change The environment and climate change, together with its implications for EU policies in a number of other areas, such as agriculture and energy, is one of the global challenges we are targeting in our 2013-2017 strategy. So our report on EU-funded waste water treatment plants in the Danube river basin examined whether EU support was helping Member States achieve its waste water policy objectives — a key element of a successful environmental policy. We concluded that while EU financial support had played an important role in bringing forward water collection and treatment, this had not been sufficient to meet the relevant deadlines. A later report in 2015 then examined a different aspect of water quality in the Danube basin: progress by the same four Member States in meeting the requirements of the water framework directive, whose main objective from 2000 was to achieve good water quality by 2015 (with some exceptions possible). We concluded that the implementation of measures co-funded by the EU had led to little improvement in water quality. Indeed, while the introduction of river basin management plans as a tool for integrated water management had been a strength of the directive, in practice the Member States’ plans suffered from shortcomings in the identification of pollution pressures and in the definition of remedial measures. On a different theme, we examined the EU’s emissions trading scheme (ETS). This is an essential pillar of the EU’s policy to combat climate change, and is designed to promote reductions of greenhouse gas emissions in a cost-effective and economically efficient way. We found that management of the scheme by the Commission and the Member States was not adequate in all respects. It was hindered by some issues related to the robustness of the framework for protecting its integrity, such as the absence of EU level oversight of the emissions market, and by significant weaknesses in the implementation of the second phase of the scheme in the period 2008-2012, after the initial preparatory phase. The Commission is committed to continue improving the EU ETS; it will address the shortcomings identified by the auditors. A number of recommendations are to be implemented by Member States. Many points noted by the ECA were specific to the period 2008-12 (phase 2) and have already largely been addressed in the current rules for the period 2013-2020 (phase 3). Since the audit report, there has been a new legislative proposal on ETS (implementation of the climate and energy objectives for 2030 following the October 2014 European Council conclusions). We also published a report about the security of energy supply and the development of the EU internal energy market. We found that the EU’s objective of completing the internal energy market by 2014 had not been reached. Energy infrastructure in Europe is generally not well designed for fully integrated markets and therefore does not currently provide effective security of energy supply. Financial support from the EU budget in the field of energy infrastructure has made only a limited contribution to the internal energy market and security of energy supply. The Commission is taking concrete steps to remove the remaining market barriers to the internal energy market, in particular through regional cooperation frameworks involving different Member States and in the context of preparation of its market design initiative. With regard to infrastructure, it is important to note that:

(a) energy infrastructure is expected to be financed by the market, i.e. user tariffs approved by independent regulators; financial support from EU budget therefore should be the exception rather than the rule;

(b) where financial support has been allocated, security of supply has often been significantly improved. A good example are the Baltic States where support from both EEPR and CEF to key projects has not only increased security of supply in electricity but also helped wholesale prices to converge.

The follow-up of special report recommendations Following up our performance audit reports is an essential element in the cycle of accountability and helps to encourage the implementation of audit recommendations by the Commission and the Member States. In our 2016 special report on the 2014 follow up exercise, we made a number of recommendations, suggesting that the Commission should bring its follow-up practices into line with relevant internal control standards. The Commission accepted these recommendations and agreed to take corrective actions. The Commission is committed to ensuring that recommendations made by the Court in its special reports and accepted by the Commission are systematically implemented and followed-up. The Commission considers that its control arrangements are in line with international control standards and best practices. However, the Commission is committed to address further challenges concerning the discharge follow-up system and procedures. THE COURT’S OBSERVATIONS We assessed the Commission’s follow up of a sample of 90 audit recommendations from 11 special reports published in the period 2011-2012. The sample was selected on the basis of a range of criteria. Of the 83 recommendations that could be verified, we concluded that the Commission fully implemented 63 % of our recommendations, 26 % were implemented in most respects, 10 % in some respects while 1 % were not implemented. Although the Commission is our principal auditee, we frequently identify matters which can best be addressed by the Member State authorities or in cooperation with the Commission, given that most EU expenditure is under shared-management arrangements with Member States. In our special reports we sometimes also address recommendations to Member States in general. We can also direct recommendations to Member States individually although this is rare. Recommendations are communicated to Member State level although formal follow-up is mixed When we make recommendations applicable to Member States in general, there are five ways in which these messages may be transmitted (see Figure 3.7). First, each report is sent to the Supreme Audit Institution (the principal national public audit body). The reports are also sent to the Council, for its consideration in the context of the discharge procedure and, on occasion, through presentations to Council working parties. Our annual activity report (which contains details of our special reports) is also sent to the parliament of each Member State, and individual Members of the Court may bring particular reports to the special attention of a national parliament or its relevant committee, in the course of their interaction with the parliament concerned. Our recommendations are also disseminated to Member States through the medium of Commission guidelines and training interventions. Communication of findings to Member States Source: European Court of Auditors. THE COURT’S OBSERVATIONS However, to date we have had relatively little information on the impact of these communication efforts in the Member States. To remedy this, we conducted a survey of paying agencies, ministries and SAIs at Member State national level on the follow up of recommendations. Four Commission DGs also completed a questionnaire. This confirmed that communication of recommendations does indeed get through to Member State level with the survey providing evidence of receipt of our messages through one or more of the above channels. In terms of impact, 80 % of respondents acknowledged at least some awareness of our reports at Member State level. However, only 13 % responded that there was a co-ordinated follow-up process for all of our recommendations in their Member State. As regards follow-up action taken, 38 % suggested that action only takes place for particular topics, while 49 % indicated that no formal follow-up action takes place. Nevertheless, some Member States provided concrete examples of embedding our recommendations in national policy and practice, mainly in the field of natural resources (see Figure 3.8). Examples of awareness and communication at Member State level Our report on The cost-effectiveness of EU Rural Development support for non-productive investments in agriculture recommended that Member States should put in place procedures to ensure that the costs of the supported non-productive investments do not exceed the costs of similar types of goods, service or works offered by the market. More concretely, Member States should define appropriate benchmarks and/or reference costs against which the costs of non-productive investments are systematically verified as part of their administrative checks. Responding to this recommendation, one Member State has introduced standard prices for the management of planting of hedges. For disseminating the Court's recommendations, a managing authority in one Member State makes the Court's findings available on its rural development programme implementation committee intranet where all the implementing bodies and the paying agency are represented. In one Member State, the Ministry of Agriculture notifies the paying agency and, where necessary, other stakeholders by means of a summary report with our key findings and recommendations, with the full report in an annex. Source: Select survey responses from our survey of SAIs and paying agencies. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The survey exercise shows that there is a reasonable level of awareness of our performance audit recommendations at Member State level. However, there is a wide variation in the level of formal follow-up and only moderate evidence of changes in national policy and practice as a result of our recommendations. It is clear that there is scope for improvement in how we develop and distribute our recommendations for maximum impact at Member State level. In the autumn of 2016, the Court will engage the Contact Committee, Commission services and the Council in the consultation on possible follow-up of recommendations in Member States. CONCLUSIONS AND RECOMMENDATIONS Conclusions Horizon 2020 is an ambitious, wide-ranging programme, bringing previously separate research and innovation programmes together for the first time, and incorporating a focus on Societal challenges. There are high-level links between Horizon 2020 and the EU’s main strategy, Europe 2020: two of the three main indicators for Horizon 2020 are also high-level indicators for Europe 2020. These indicators, however, are of limited use in tracking the contribution of Horizon 2020 to Europe 2020 (see paragraphs 3.7 to 3.17). While the Commission accepts that it needs to improve further, it considers that its performance framework already tracks the contribution of Horizon 2020 to Europe 2020. As explained under Figure 3.1 and in paragraph 3.13: The EU budget represents only a small part of total research and innovation spending in Europe, and so in itself can make only a minor contribution towards the achievement of the Europe 2020 objectives. Progress towards the Europe 2020 strategy depends on a wide range of actions at EU and national level, which are not only on spending, and for which Member States have a primordial role to play. The contribution of Horizon 2020 to the achievement of the Europe 2020 strategy is reinforced by the interaction with other actions at EU and national level. The links between the Commission’s 10 political priorities and Europe 2020/Horizon 2020 need further clarification. The political priorities overlap with, but are different from, the Europe 2020 priorities. It is not clear to which set of priorities Horizon 2020 — in its planning right through to its monitoring and reporting — is now to be aligned. There is a risk that other areas of EU activity are concerned (see paragraphs 3.18 to 3.21). The political priorities of the Juncker Commission provide the framework for the Commission’s own contribution to the delivery of the Europe 2020 strategy, which is a shared responsibility of the Union and the Member States. The political priorities guide the Commission’s implementation of financial programmes but this is done in full complementarity with the Europe 2020 strategy and in accordance with the legal framework underpinning Horizon 2020, which has not changed. A key success factor for Horizon 2020 is effective synergy and complementarity between national and European research and innovation programmes. Our partners in the supreme audit institutions of Bulgaria and Portugal found that while there were areas of complementarity in their countries, there were also some problems at the national level (see paragraphs 3.22 and 3.23). The legal framework of Horizon 2020 introduces several important elements for performance management such as objectives and key performance indicators. This is a welcome development. However, further developments to the framework are still needed; some baselines, milestones and targets still need to be developed, some targets should be reviewed to ensure that they are suitably ambitious; some indicators suffer from lack of available information. The objectives defined by the legislation are themselves naturally at a high-level, but are not translated, through work programmes and calls, into lower level objectives useful for operational performance management purposes. The Commission does not use the Horizon 2020 work programmes and associated calls to increase the focus on performance (see paragraphs 3.27 to 3.38). The performance framework was discussed with the legislative authority during the adoption of the Horizon 2020 legislation, and is set out in that legislation. It includes a number of indicators that will be monitored throughout the programme, and interim and ex post evaluations. This clear framework is an advance on previous framework programmes. The results and impacts of work programmes, and of Horizon 2020 more generally, will be assessed using this performance framework. The long-term nature of research, the timescale of the preparation of the legislation, and the limitations in the availability of some data do provide challenges in the completion of the performance framework, and the Commission accepts that continual improvements are necessary. For the proposals and grant agreements we examined, objectives were generally SMART when it was required by the Commission. The evaluation process of these proposals, with the exception of ERC that uses excellence as a criterion, had an emphasis on performance. Grant agreements require beneficiaries to report to the Commission information for the aggregation of the Horizon 2020 KPIs. However the use of the wider concept of expected impact rather than expected result increases the risk that information provided for this part is too broad and not directly attributable to the action funded by Horizon 2020, with the result that performance assessment could be difficult (see paragraphs 3.39 to 3.43). The use of expected impact rather than result arises from the fact that the Horizon 2020 legislation predates the better regulation guidelines. The Commission wants to evaluate the expected impact of projects when considering which projects to support. This will not affect the monitoring of the indicators that have been set out in the legislation, and which are the cornerstone of the performance framework. Weaknesses in the design of the monitoring and reporting system for Horizon 2020 mean that the Commission is limited in its ability to monitor and report on the performance of the programme: The Commission considers that the Horizon 2020 monitoring and reporting system is a great improvement compared to FP7. Clear objectives have been established, and a performance framework, discussed fully with stakeholders and the legislative authority, has been anchored in the legal basis. The Commission is aware of some weaknesses that remain, and is working to remedy them:

(a) The current setup does not enable the Commission to monitor and report separately the spending and the performance of the different components of research and development and innovation within Horizon 2020 (see paragraphs 3.45 to 3.47).

(a) the Commission does not consider that research and innovation spending needs to be reported on separately, this is not required by the legislation and runs counter to the efforts made in Horizon 2020 to close the gap between research and innovation;

(b) The contribution of Horizon 2020 to Europe 2020 is well established in the budgetary process. However, thus far the contribution of Horizon 2020 to Europe 2020 has not yet been reported on in a meaningful way (see paragraphs 3.48 to 3.50).

(b) the legislation requires the contribution of Horizon 2020 to Europe 2020 to be reported on in the interim evaluation of the programme, and this evaluation will be undertaken in 2017. The nature of the research cycle means that a certain time must elapse before the contribution to Europe 2020 can be reported on in a meaningful way;

(c) The Horizon 2020 monitoring and reporting system has improved compared to FP7 but there remain weaknesses regarding the information gathered for Horizon 2020. There is a risk that difficulties encountered in the ex post evaluation of FP 7 might persist for Horizon 2020 (see paragraphs 3.51 to 3.53).

(c) the Commission agrees that the Horizon 2020 monitoring and reporting system has improved compared to FP7. Clear objectives have been established, and a performance framework, discussed fully with stakeholders and the legislative authority, has been anchored in the legal basis.

It considers that it has acted to mitigate the difficulties identified in the FP7 evaluation system. In assessing performance, some basic concepts are the different types of benefits which can result from public action: inputs, outputs, results and impacts. While the Commission has defined how these terms should be used, it does not always abide by these definitions. A particular problem is the apparently arbitrary use of result and impact in translating the legislation into individual work programmes (see paragraphs 3.54 to 3.59). The adoption of the Horizon 2020 Regulations (December 2013) preceded the better regulation guidelines (May 2015). This has led to some inconsistencies, as noted by the Court. Since the adoption of the better regulation guidance, the better regulation terminology is consistently used in the monitoring and evaluation of the performance of Horizon 2020. For the natural resources family of DGs, we noted that many of the objectives used in MPs and AARs were taken directly from policy or legislative documents and were therefore at too high a level for management purposes. We observed progress for the indicators used to measure performance but some shortcomings remain (see paragraphs 3.60 to 3.65). Since the Court’s observation in relation to the management plans for 2015, the Commission has introduced changes to the SPP cycle notably to provide a clearer framework for the Commission’s accountability and for its services. With the new approach, the planning documents as from 2016 are meant to be more streamlined and focused on the priorities of the Commission and the competencies of the DGs, helping them to establish priorities and tell a coherent performance story with a clear distinction between attribution (for results and outputs that are a direct consequence of the Commission services’ activities) and contribution (where the Commission services’ activities play a part in bringing about a given result or an impact) for EU high level policy and spending programmes’ results. The strategic plan introduces a new approach shifting the focus from short-term planning to a long-term, strategic perspective helping the DGs to align their specific objectives with the long-term objectives defined at Commission. The operational planned outputs for the year ahead are listed in the Management Plan. Recommendations Annex 3.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, we presented five recommendations. Of these recommendations, two were implemented in some respects and three were not implemented. Our recommendations for 2015 are as follows:

Recommendation 1: Operational objectives The Commission should translate the Horizon 2020 legislation high-level objectives into operational objectives at work programme level, so that, by assessing the performance of work programmes and calls, they can effectively be used as drivers for performance.

The Commission accepts this recommendation. While the overall performance framework is laid down and clear, the Commission notes the points raised by the Court, which will be fully taken into account as part of work underway to improve the work programmes, including its role in performance based reporting.

Recommendation 2: Further clarification for the different strategies are needed The Commission should further clarify the links between the Europe 2020 strategy (2010-2020), the multi-annual financial framework (2014-2020), the Commission priorities (2015-2019) through, for instance, the strategic planning and reporting process (2016-2020). This would strengthen monitoring and reporting arrangements and enable the Commission to report effectively on the contribution of the EU budget towards Europe 2020 objectives.

The Commission accepts this recommendation. The political priorities of the Commission are clearly defined in President Juncker’s political guidelines. They provide a roadmap for the Commission’s action that is fully consistent and compatible with Europe 2020 as Europe’s long-term growth strategy. The differences in scope between the priorities and Europe 2020 reflect the Commission’s duty to respond to challenges that have emerged or grown in prominence since the development of the Europe 2020 strategy, such as migration. From 2016 onwards and based on the strategic plans produced by each Commission department, the annual activity reports of each Commission department as well as the annual management and performance report of the Commission will explain how specific objectives contribute to the political priorities and will further clarify the links with Europe 2020.

Recommendation 3: Consistent terminology use The Commission should, across all its activities, use the terms input, output, result and impact consistently and in line with its better regulation guidelines.

The Commission accepts this recommendation. The adoption of the different Regulations for the 2014-2020 MFF period preceded the better regulation guidelines (May 2015). Moreover, the content of those regulations is the result of a political negotiation between the co-legislators. This has led to some inconsistencies, as noted by the Court. Since the adoption of the better regulation guidance, the better regulation terminology should be consistently used in the monitoring and evaluation of all programmes. Consistent use of terms across all legislation may, however, not be possible in the short term. With regard to the Commission’s internal performance framework, the Commission has provided consistent definitions and methodologies through the standing instructions provided to all departments. These definitions are equivalent to those contained in the better regulation guidelines.

ANNEX

(1) No 1/2015 Inland Waterway Transport in Europe: No significant improvements in modal share and navigability conditions since 2001 (2) No 2/2015 EU-funding of Urban Waste Water Treatment plants in the Danube river basin: further efforts needed in helping Member States to achieve EU waste water policy objectives (3) No 3/2015 EU Youth Guarantee: first steps taken but implementation risk ahead (4) No 4/2015 Technical assistance: what contribution has it made to agriculture and rural development? (5) No 5/2015 Are financial instruments a successful and promising tool in the rural development area? (6) No 6/2015 The integrity and implementation of the EU Emissions Trading Scheme (7) No 7/2015 The EU police mission in Afghanistan: mixed results (8) No 8/2015 Is EU financial support adequately addressing the needs of micro-entrepreneurs? (9) No 9/2015 EU support for the fight against torture and the abolition of the death penalty (10) No 10/2015 Efforts to address problems with public procurement in EU cohesion expenditure should be intensified (11) No 11/2015 Are the Fisheries Partnership Agreements well managed by the Commission? (12) No 12/2015 The EU priority of promoting a knowledge-based rural economy has been affected by poor management of knowledge-transfer and advisory measures (13) No 13/2015 EU support to timber-producing countries under the FLEGT action plan (14) No 14/2015 The ACP Investment Facility: does it provide added value? (15) No 15/2015 ACP-EU Energy Facility support for renewable energy in East Africa (16) No 16/2015 Improving the security of energy supply by developing the internal energy market: more efforts needed (17) No 17/2015 Commission’s support of youth action teams: redirection of ESF funding achieved, but insufficient focus on results (18) No 18/2015 Financial assistance provided to countries in difficulties (19) No 19/2015 More attention to results needed to improve the delivery of technical assistance to Greece (20) No 20/2015 The cost-effectiveness of EU Rural Development support for non-productive investments in agriculture (21) No 21/2015 Review of the risks related to a results-oriented approach for EU development and cooperation action (22) No 22/2015 EU supervision of credit rating agencies — well established but not yet fully effective (23) No 23/2015 Water quality in the Danube river basin: progress in implementing the water framework directive but still some way to go (24) No 24/2015 Tackling intra-Community VAT fraud: More action needed (25) No 25/2015 EU support for rural infrastructure: potential to achieve significantly greater value for money

No 1/2015 Inland Waterway Transport in Europe: No significant improvements in modal share and navigability conditions since 2001

No 2/2015 EU-funding of Urban Waste Water Treatment plants in the Danube river basin: further efforts needed in helping Member States to achieve EU waste water policy objectives

No 3/2015 EU Youth Guarantee: first steps taken but implementation risk ahead

No 4/2015 Technical assistance: what contribution has it made to agriculture and rural development?

No 5/2015 Are financial instruments a successful and promising tool in the rural development area?

No 6/2015 The integrity and implementation of the EU Emissions Trading Scheme

No 7/2015 The EU police mission in Afghanistan: mixed results

No 8/2015 Is EU financial support adequately addressing the needs of micro-entrepreneurs?

No 9/2015 EU support for the fight against torture and the abolition of the death penalty

No 10/2015 Efforts to address problems with public procurement in EU cohesion expenditure should be intensified

No 11/2015 Are the Fisheries Partnership Agreements well managed by the Commission?

No 12/2015 The EU priority of promoting a knowledge-based rural economy has been affected by poor management of knowledge-transfer and advisory measures

No 13/2015 EU support to timber-producing countries under the FLEGT action plan

No 14/2015 The ACP Investment Facility: does it provide added value?

No 15/2015 ACP-EU Energy Facility support for renewable energy in East Africa

No 16/2015 Improving the security of energy supply by developing the internal energy market: more efforts needed

No 17/2015 Commission’s support of youth action teams: redirection of ESF funding achieved, but insufficient focus on results

No 18/2015 Financial assistance provided to countries in difficulties

No 19/2015 More attention to results needed to improve the delivery of technical assistance to Greece

No 20/2015 The cost-effectiveness of EU Rural Development support for non-productive investments in agriculture

No 21/2015 Review of the risks related to a results-oriented approach for EU development and cooperation action

No 22/2015 EU supervision of credit rating agencies — well established but not yet fully effective

No 23/2015 Water quality in the Danube river basin: progress in implementing the water framework directive but still some way to go

No 24/2015 Tackling intra-Community VAT fraud: More action needed

No 25/2015 EU support for rural infrastructure: potential to achieve significantly greater value for money

Anx. 32. Year Court recommendation Court's analysis of the progress made Commission reply Fully implemented Being implemented Not implemented Not applicable Insufficient evidence In most respects In some respects 2013 Recommendation 1: Following this review and the findings and conclusions for 2013, the Court recommends that the Commission should, on the next occasion that the Financial Regulation is reviewed, rationalise its reporting framework for performance. X The Commission underlines that the review is still ongoing. Recommendation 2: Following this review and the findings and conclusions for 2013, the Court recommends that the Commission should ensure that the evaluation report presents a summary account that brings together all the information available on the progress towards Europe 2020 targets in order to provide the reader with a clear overview of the achievements made. X The Commission is continuously aiming to improve its reporting on the progress towards Europe 2020 targets in the annual management and performance report as well as in the context of the EU semester. Recommendation 3: Following this review and the findings and conclusions for 2013, the Court recommends that the Commission should further develop its performance managing and reporting system so that it allows the Commission to take responsibility for sound financial management as well as the EU budget’s contribution to policy achievements in the annual declarations of assurance by the directors-general. X The Commission did not accept this recommendation. 2012 Recommendation 1: The Commission and the legislator should ensure that there is a focus on performance in the forthcoming programming period (2014-2020). This requires that a limited number of sufficiently specific objectives with relevant indicators, expected results and impacts are laid down in the sector-specific regulations or in some other binding manner. X The Commission considers that the focus on performance has been considerably strengthened in the 2014-2020 legislative framework. Recommendation 2: Commission should ensure that there is a clear link between the DGs' activities and the objectives set. When identifying these objectives each DG should take into account the relevant management mode, where applicable, and its role and responsibilities. X The Commission did not accept this recommendation.

ANNEX 3.2 FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR PERFORMANCE ISSUES

Year Court recommendation Court's analysis of the progress made Commission reply Fully implemented Being implemented Not implemented Not applicable Insufficient evidence In most respects In some respects 2013 Recommendation 1: Following this review and the findings and conclusions for 2013, the Court recommends that the Commission should, on the next occasion that the Financial Regulation is reviewed, rationalise its reporting framework for performance. X The Commission underlines that the review is still ongoing. Recommendation 2: Following this review and the findings and conclusions for 2013, the Court recommends that the Commission should ensure that the evaluation report presents a summary account that brings together all the information available on the progress towards Europe 2020 targets in order to provide the reader with a clear overview of the achievements made. X The Commission is continuously aiming to improve its reporting on the progress towards Europe 2020 targets in the annual management and performance report as well as in the context of the EU semester. Recommendation 3: Following this review and the findings and conclusions for 2013, the Court recommends that the Commission should further develop its performance managing and reporting system so that it allows the Commission to take responsibility for sound financial management as well as the EU budget’s contribution to policy achievements in the annual declarations of assurance by the directors-general. X The Commission did not accept this recommendation. 2012 Recommendation 1: The Commission and the legislator should ensure that there is a focus on performance in the forthcoming programming period (2014-2020). This requires that a limited number of sufficiently specific objectives with relevant indicators, expected results and impacts are laid down in the sector-specific regulations or in some other binding manner. X The Commission considers that the focus on performance has been considerably strengthened in the 2014-2020 legislative framework. Recommendation 2: Commission should ensure that there is a clear link between the DGs' activities and the objectives set. When identifying these objectives each DG should take into account the relevant management mode, where applicable, and its role and responsibilities. X The Commission did not accept this recommendation.

CHAPTER 4 Revenue

Txt. 1. THE COURT’S OBSERVATIONS INTRODUCTION This chapter presents our findings for revenue, which consists of own resources and other revenue. Key information on revenue in 2015 is provided in Figure 4.1. Revenue — Key information 2015 (billion euro) Source: 2015 consolidated accounts of the European Union. THE COURT’S OBSERVATIONS Brief description of revenue Most revenue comes from own resources (90 %) i.e. contributions from Member States, of which there are three categories: (5) The gross national income (GNI)-based own resource (64 % of revenue) results from the application of a uniform rate to the Member States' GNI. The calculation of Member States' contributions is based on forecast GNI data. After taking into account all other sources of revenue, the GNI-based own resource is used to balance the EU budget. The principal risks are that the underlying statistics are not compiled in compliance with EU rules or are not processed by the Commission according to these rules in the calculation of the Member States’ contributions, including balances and adjustments. Another risk is that the Commission’s checks on the Member States’ GNI data do not detect potential non-compliance with EU rules. (6) The value added tax (VAT)-based own resource (13 %) is derived from the application of a uniform rate to Member States’ notionally harmonised VAT assessment bases. The main risks lie in the completeness and accuracy of the information provided by Member States, the accuracy of the Commission’s calculations of the contributions due and the timeliness of the Member States’ payments. (7) Traditional own resources (TOR) (13 %) are customs duties collected on imports and sugar production charges. They are established and collected by the Member States. Three quarters are paid into the EU budget, the remaining quarter being retained to cover collection costs. The principal risks regarding TOR are the completeness, accuracy and timeliness of the duties made available to the EU. Txt. 2. The main risks in respect of other revenue include the Commission’s management of fines, and errors in the calculation of contributions relating to EU and other agreements and correction mechanisms for budgetary imbalances. On 26 May 2014, a new own resources legislative package was formally approved by the Council. When ratified by the Member States, it will be applied with retroactive effect from 1 January 2014. We will follow up the ratification process and the subsequent calculation of the retroactive impact. Audit scope and approach Annex 1.1 sets out the key elements of our audit approach and methodology. For the audit of revenue, the following specific points should be noted: (i) the Commission’s systems for ensuring that the Member States’ GNI data is appropriate as a basis for own-resources purposes (including for the lifting of reservations), and its systems for calculating and collecting the GNI-based own resource contributions and GNI balances; (ii) the Commission’s systems for ensuring that VAT-based own resources and VAT balances are correctly calculated and collected; (iii) the Commission’s systems for TOR, including its monitoring of the Member States’ post-clearance audits; (iv) the TOR accounting systems in three selected Member States (France, Italy and Lithuania) and a review of their systems for post-clearance audits; (v) the Commission's management of fines and penalties under Title 7 of the EU budget; (vi) the Commission’s calculation of the European Economic Area (EEA)/European Free Trade Association (EFTA) contributions; (vii) the Commission’s calculation of the correction mechanisms. (a) We examined the following systems: (b) We examined a sample of 55 Commission recovery orders. The sample is designed to be representative of all sources of revenue. (c) We assessed the annual activity reports of the Commission’s Directorate-General for Budget (DG Budget) and Eurostat. Txt. 3. REGULARITY OF TRANSACTIONS Annex 4.1 contains a summary of the results of transaction testing. Of the 55 transactions examined, none was affected by error. (11) We found no errors in the Commission’s calculation of Member States' contributions on the basis of the VAT and GNI data or their payment. (12) We found that the Commission’s recovery orders reflected the TOR statements sent by the Member States. (13) We found no errors in the calculations or payments of other revenue transactions. Txt. 4. EXAMINATION OF ANNUAL ACTIVITY REPORTS AND OTHER ELEMENTS OF INTERNAL CONTROL SYSTEMS As explained in paragraph 4.5, we examined selected systems. This section provides detailed observations that do not qualify our view on revenue, but lead to recommendations to further improve the calculation and collection of revenue. GNI and VAT balances and adjustments reflect updates of Member States GNI and VAT figures and work on reservations Every year Member States have to provide updates of their GNI and VAT data for the years still subject to change (four years) or to address the points under reservation. The Commission uses this information to recalculate Member States’ GNI and VAT contributions for the previous years (VAT and GNI balances and adjustments). A substantial part of the significant corrections to Member States contributions calculated in 2014 (9813 million euro) resulted from the work on GNI reservations, which led the Commission to lift most of those reservations during 2015. The Commission lifted a large number of GNI reservations GNI reservations can be general or specific. General reservations cover all elements of the GNI compilation. Specific reservations cover discrete elements of GNI and can be transaction-specific, transversal-specific or process-specific. Figure 4.2 shows the evolution of the number of reservations in place at the year end. It also includes information related to TOR and VAT discussed later in this chapter. Detailed information on the situation at the end of 2015 by Member State can be found in Annex 4.3. Evolution of the number of GNI reservations, VAT reservations and TOR open points at the year end The GNI process-specific (see paragraph 4.11) and general reservations (see paragraph 4.12) are not included in the GNI total. Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES There was a substantial reduction in the number of transaction-specific and transversal-specific reservations in place In the course of 2015, the Commission lifted 184 reservations (67 transaction-specific and 117 transversal-specific), including the long outstanding transaction-specific GNP reservation on Greece, relating to the 1995-2001 period. We did not identify serious problems in the sample of lifted reservations that we analysed. At the end of 2015, there were 55 reservations covering discrete elements of the Member States’ compilations (see Figure 4.2), of which 39 were transaction-specific and 16 transversal-specific. The Commission attaches a high priority to the verification of GNI and lifting the reservations. Consequently, the GNI team at Eurostat was reinforced by a temporary redeployment of staff in the first half of 2015. Subsequently, the personnel was almost doubled towards the end of the year. The number of outstanding reservations was reduced still further during the early part of 2016. THE COURT’S OBSERVATIONS Process specific reservations were extended The Commission extended the 27 process-specific reservations on the Member States’ 2010 GNI compilation in order to cover the year 2011 — these were the only reservations set in 2015. They were set in order to allow the Commission to perform a check on the data established according to the European System of National and Regional Accounts (ESA 2010). There are still general reservations on Greek GNI data covering 2008 and 2009 The only general reservations in place at the end of 2015 were on Greek GNI data for the years 2008 and 2009. Although progress had been made in addressing these reservations, they were not lifted. However, during 2015, the Commission continued to follow closely the problems with Greece’s compilation of national accounts and kept the technical assistance programme that had been set up in 2014 with the objective of creating an independent and stronger national statistical authority. This work improves the reliability of the Greek GNI data. It is important, as the quality of that data can have an impact on all Member States’ individual contributions in terms of GNI-based own resources (see footnote 4). The number of VAT-based own resources reservations has decreased At the end of the year, there were a total of 85 reservations, compared to 101 in 2014 (see Figure 4.2), of which 69 had been set by the Commission and 16 by the Member States. Of the 69 reservations placed by the Commission, 15 related to infringements of the VAT directive (such as granting exemptions to products or activities that should not be exempted). The number of long-outstanding reservations set by the Commission decreased from ten to six. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The management of traditional own resources by Member States can still be improved Post-clearance audits are an important tool to safeguard the financial interests of the EU,… Customs authorities may check the commercial documents and data related to goods after they have been imported and released for free circulation in the EU. These checks, which include post-clearance audits, are referred to as post-clearance controls. They should be based on risk analyses using criteria developed at the national, EU and, where available, international level. The Commission will continue to work with Member States in line with the Commission Strategy and Action Plan for customs risk management to ensure controls (including post-clearance controls) are effective and risk-based, notably through the further development of EU common risk criteria and standards for financial risks. … but lack of harmonization persists, affecting the results As reported in previous years’ annual reports, we found weaknesses in the identification, selection and inspection of the importers audited by Member States. As an example, in Italy, the selection of post-clearance audits is typically based on the characteristics of individual transactions and not, as recommended in the Commission’s customs audit guide, on the profile of companies. The Commission, in its inspections, regularly examines the post-clearance controls including audits carried out by the Member States and encourages them to make use of the Customs Audit Guide in the conduct of those controls/audits. It will follow up the findings in its regular follow-up of the findings resulting from the Court’s audits and its own inspections. See Commission reply to paragraph 4.14. The quality, scope and results of the post-clearance audits varied substantially across the three Member States visited. A specific case is France, where, for post-clearance audits, a debt can be notified more than three years after it was incurred. This approach differs from that in other Member States because the audits are, according to French legislation, always linked to the prevention and pursuit of customs law infringements and are thus liable to give rise to criminal proceedings. Therefore, the French authorities interrupt the three year time-barring period during post-clearance audits. While there may be acceptable grounds for difference in the scope and results of post-clearance audits carried out across the Member States, the Commission in its inspections in Member States will continue to examine the points raised by the Court concerning the post-clearance audits carried out by the national authorities. The Member States have access to the Customs Audit Guide, the use of which is not mandatory but which should guide them in the conduct of the post-clearance controls they carry out. In accordance with EU customs legislation, a three-year time limit applies for the notification of a customs debt except where the case may give rise to criminal proceedings. The determination of whether a case may give rise to such proceedings differs across the Member States. The proposed Directive on the Union legal framework for customs infringements and sanctions (COM(2013) 884 final), once adopted and applied, is expected to address the issue of systematic use of Article 221(4) of the Community Customs Code and now Article 103(2) Union Customs Code by some Member States. The management of A and B accounts still needs to be improved in some Member States Each Member State sends the Commission a statement of its customs duties and sugar levies established on a monthly basis (the A accounts) and a quarterly statement of established duties which are not included in the A accounts (the B accounts). The Commission detected shortcomings in the Member States’ management of the B accounts in 17 out of 22 Member States visited. As in previous years, we also found inefficiencies in the management of the B accounts. We noted good practices in Lithuania, even though debt recovery was not always successful (see Figure 4.3). In Italy some systematic weaknesses need to be addressed. The Italian authorities could not confirm that all the amounts collected from the B accounts were included in the correct A account statement. The Commission identified two other Member States (Belgium and France) with similar inconsistencies between the A and B accounts statements. These shortcomings did not affect the reliability of the overall amounts as disclosed in the consolidated accounts of the European Union. The Commission examines systematically the management of the B account in each of its inspections in the Member States, and when it finds shortcomings, requests the Member States to rectify them. In the Advisory Committee on the Communities’ Own Resources (ACOR) of December 2015, the Commission services raised again the problem concerning the consistency between the A and B accounts and requested the Member States whose A and В account statements continued to show inconsistencies to supply a plan to resolve them containing a timeline within which they will be resolved and to carry out routine checks to prevent a recurrence of these inconsistencies. For debtor companies registered outside the EU or debts of citizens of countries outside the EU, there is no effective instrument for debt recovery. In six out of 24 cases sampled in Lithuania, the debtors were from non EU countries: citizens of Russia, Belarus or Ukraine and companies registered in Turkey and the British Virgin Islands. Even though the Lithuanian customs authorities carried out all possible recovery procedures within the borders of the EU, in the cases analysed, the debt recovery was not successful. Moreover, in four of the 15 write-off cases that we selected for analysis across all Member States, the authorities had not been able to enforce debt collection due to the fact that the debtors were located in Switzerland, Ukraine or Russia. The Commission considers that the entry into force of the UCC on 1 May 2016 has provided additional instruments to ensure debt collection in the situations examined by the Court. In particular, it expands the use of mandatory guarantees and it provides for optional guarantees and the appointment of jointly liable customs representatives based in the EU. The number of open points remains significant The Commission’s inspections in Member States may result in the national authorities having to take steps because of non-compliance with EU customs rules. These are known as open points. These open points can have a potential financial impact. At the end of 2015 a total of 325 points (see Figure 4.2) were open. The Commission services follow up open points in a timely and prompt manner. They will only close a point when they are satisfied that adequate measures to resolve it have been taken and when any traditional own resources due have been made available, with interest where applicable. However, complex points (e.g. requiring a change of national legislation or instructions, organisational or IT changes in national services or infringement proceedings) take time to resolve. Of the 325 points shown by the Court as open at the end of 2015, only 199 have a potential financial impact. EEA/EFTA contributions and correction mechanisms calculations contained minor errors We identified minor errors in two areas: 4.20. Pt_a_. 2. the 2015 Commission’s calculation of the EEA/EFTA contributions, and Pt_b_. 2. the 2015 Commission’s calculation of correction mechanisms. Pt_b_. 3. The Commission has taken measures to improve its checks and to avoid errors identified by the Court. For example, new checklists were introduced, procedures for preparation of inputs improved and the double checking of documents further strengthened, especially with regards to inputs used. Moreover, a more detailed description of the calculations providing better guidance for new staff involved was developed. Txt. 5. The errors were the result of incorrect calculation formula and inputs from outdated databases. These errors are not material, but they indicate that the current systems for performing and reviewing these calculations should be improved. Annual activity reports provide a fair assessment The 2015 annual activity reports by DG Budget and Eurostat provide a fair assessment of financial management in relation to the legality and regularity of underlying transactions concerning own resources and other revenue. The information provided corroborates our observations and conclusions. CONCLUSION AND RECOMMENDATIONS The conclusion for 2015 Overall audit evidence indicates that revenue is not affected by a material level of error. In particular for revenue: (17) we assessed the examined systems as effective for GNI and VAT-based own resources; (18) we assessed the examined systems as overall effective for TOR. The key internal controls in Member States visited are assessed as partially effective; (19) we found no errors in the transactions we tested. Txt. 6. Recommendations Annex 4.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, we presented seven recommendations. Of these recommendations, four were implemented in most respects and three were implemented in some respects. Following this review and the findings and conclusions for 2015, we recommend that the Commission: Traditional own resources (20) Recommendation 1: take the necessary steps to ensure that economic operators receive a similar treatment in all Member States as regards the time-barring of the debt notifications following a post-clearance audit; Txt. 7. The Commission partially accepts the recommendation. The Commission is of the view that its proposal for a Directive on the Union legal framework for customs infringements and sanctions (COM(2013) 884 final) addresses the observations of the Court and that the situation is expected to improve once the proposal is adopted and applied. However, the outcome depends on the decision of the legislative authority. (21) Recommendation 2: provide Member States with guidance to improve their management of the items recorded in the B accounts; Txt. 8. The Commission accepts the recommendation. Its services, in their inspections in the Member States, regularly provide guidance to the national authorities concerning the management of the B account and promote best practice. (22) Recommendation 3: ensure that Member States correctly declare and make available the amounts collected from the B accounts; Txt. 9. The Commission accepts the recommendation. It will continue to ensure that the Member States correctly declare and make available the amounts recovered from the B accounts. (23) Recommendation 4: facilitate to the extent possible the recovery of customs debts by the Member States, where the debtors are not based in an EU Member State. Txt. 10. The Commission accepts the recommendation. The Commission will in due course examine and assess the effectiveness of the new UCC in addressing the issues identified by the Court. See also the Commission reply to Figure 4.3. EEA/EFTA contributions and correction mechanisms (24) Recommendation 5: improve the checks on the calculations of the EEA/EFTA contributions and correction mechanisms. Txt. 11. The Commission accepts the recommendation. In order to improve the checks, it has already improved and strengthened the procedures and checklists.

THE COURT’S OBSERVATIONS INTRODUCTION This chapter presents our findings for revenue, which consists of own resources and other revenue. Key information on revenue in 2015 is provided in Figure 4.1. Revenue — Key information 2015 (billion euro) Source: 2015 consolidated accounts of the European Union. THE COURT’S OBSERVATIONS Brief description of revenue Most revenue comes from own resources (90 %) i.e. contributions from Member States, of which there are three categories:

The gross national income (GNI)-based own resource (64 % of revenue) results from the application of a uniform rate to the Member States' GNI. The calculation of Member States' contributions is based on forecast GNI data. After taking into account all other sources of revenue, the GNI-based own resource is used to balance the EU budget. The principal risks are that the underlying statistics are not compiled in compliance with EU rules or are not processed by the Commission according to these rules in the calculation of the Member States’ contributions, including balances and adjustments. Another risk is that the Commission’s checks on the Member States’ GNI data do not detect potential non-compliance with EU rules.

The value added tax (VAT)-based own resource (13 %) is derived from the application of a uniform rate to Member States’ notionally harmonised VAT assessment bases. The main risks lie in the completeness and accuracy of the information provided by Member States, the accuracy of the Commission’s calculations of the contributions due and the timeliness of the Member States’ payments.

Traditional own resources (TOR) (13 %) are customs duties collected on imports and sugar production charges. They are established and collected by the Member States. Three quarters are paid into the EU budget, the remaining quarter being retained to cover collection costs. The principal risks regarding TOR are the completeness, accuracy and timeliness of the duties made available to the EU.

The main risks in respect of other revenue include the Commission’s management of fines, and errors in the calculation of contributions relating to EU and other agreements and correction mechanisms for budgetary imbalances. On 26 May 2014, a new own resources legislative package was formally approved by the Council. When ratified by the Member States, it will be applied with retroactive effect from 1 January 2014. We will follow up the ratification process and the subsequent calculation of the retroactive impact. Audit scope and approach Annex 1.1 sets out the key elements of our audit approach and methodology. For the audit of revenue, the following specific points should be noted:

(a) We examined the following systems: (i) the Commission’s systems for ensuring that the Member States’ GNI data is appropriate as a basis for own-resources purposes (including for the lifting of reservations), and its systems for calculating and collecting the GNI-based own resource contributions and GNI balances; (ii) the Commission’s systems for ensuring that VAT-based own resources and VAT balances are correctly calculated and collected; (iii) the Commission’s systems for TOR, including its monitoring of the Member States’ post-clearance audits; (iv) the TOR accounting systems in three selected Member States (France, Italy and Lithuania) and a review of their systems for post-clearance audits; (v) the Commission's management of fines and penalties under Title 7 of the EU budget; (vi) the Commission’s calculation of the European Economic Area (EEA)/European Free Trade Association (EFTA) contributions; (vii) the Commission’s calculation of the correction mechanisms.

(a)

(i) the Commission’s systems for ensuring that the Member States’ GNI data is appropriate as a basis for own-resources purposes (including for the lifting of reservations), and its systems for calculating and collecting the GNI-based own resource contributions and GNI balances;

(ii) the Commission’s systems for ensuring that VAT-based own resources and VAT balances are correctly calculated and collected;

(iii) the Commission’s systems for TOR, including its monitoring of the Member States’ post-clearance audits;

(iv) the TOR accounting systems in three selected Member States (France, Italy and Lithuania) and a review of their systems for post-clearance audits;

(v) the Commission's management of fines and penalties under Title 7 of the EU budget;

(vi) the Commission’s calculation of the European Economic Area (EEA)/European Free Trade Association (EFTA) contributions;

(vii) the Commission’s calculation of the correction mechanisms.

(b) We examined a sample of 55 Commission recovery orders. The sample is designed to be representative of all sources of revenue.

(c) We assessed the annual activity reports of the Commission’s Directorate-General for Budget (DG Budget) and Eurostat.

REGULARITY OF TRANSACTIONS Annex 4.1 contains a summary of the results of transaction testing. Of the 55 transactions examined, none was affected by error.

We found no errors in the Commission’s calculation of Member States' contributions on the basis of the VAT and GNI data or their payment.

We found that the Commission’s recovery orders reflected the TOR statements sent by the Member States.

We found no errors in the calculations or payments of other revenue transactions.

EXAMINATION OF ANNUAL ACTIVITY REPORTS AND OTHER ELEMENTS OF INTERNAL CONTROL SYSTEMS As explained in paragraph 4.5, we examined selected systems. This section provides detailed observations that do not qualify our view on revenue, but lead to recommendations to further improve the calculation and collection of revenue. GNI and VAT balances and adjustments reflect updates of Member States GNI and VAT figures and work on reservations Every year Member States have to provide updates of their GNI and VAT data for the years still subject to change (four years) or to address the points under reservation. The Commission uses this information to recalculate Member States’ GNI and VAT contributions for the previous years (VAT and GNI balances and adjustments). A substantial part of the significant corrections to Member States contributions calculated in 2014 (9813 million euro) resulted from the work on GNI reservations, which led the Commission to lift most of those reservations during 2015. The Commission lifted a large number of GNI reservations GNI reservations can be general or specific. General reservations cover all elements of the GNI compilation. Specific reservations cover discrete elements of GNI and can be transaction-specific, transversal-specific or process-specific. Figure 4.2 shows the evolution of the number of reservations in place at the year end. It also includes information related to TOR and VAT discussed later in this chapter. Detailed information on the situation at the end of 2015 by Member State can be found in Annex 4.3. Evolution of the number of GNI reservations, VAT reservations and TOR open points at the year end The GNI process-specific (see paragraph 4.11) and general reservations (see paragraph 4.12) are not included in the GNI total. Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES There was a substantial reduction in the number of transaction-specific and transversal-specific reservations in place In the course of 2015, the Commission lifted 184 reservations (67 transaction-specific and 117 transversal-specific), including the long outstanding transaction-specific GNP reservation on Greece, relating to the 1995-2001 period. We did not identify serious problems in the sample of lifted reservations that we analysed. At the end of 2015, there were 55 reservations covering discrete elements of the Member States’ compilations (see Figure 4.2), of which 39 were transaction-specific and 16 transversal-specific. The Commission attaches a high priority to the verification of GNI and lifting the reservations. Consequently, the GNI team at Eurostat was reinforced by a temporary redeployment of staff in the first half of 2015. Subsequently, the personnel was almost doubled towards the end of the year. The number of outstanding reservations was reduced still further during the early part of 2016. THE COURT’S OBSERVATIONS Process specific reservations were extended The Commission extended the 27 process-specific reservations on the Member States’ 2010 GNI compilation in order to cover the year 2011 — these were the only reservations set in 2015. They were set in order to allow the Commission to perform a check on the data established according to the European System of National and Regional Accounts (ESA 2010). There are still general reservations on Greek GNI data covering 2008 and 2009 The only general reservations in place at the end of 2015 were on Greek GNI data for the years 2008 and 2009. Although progress had been made in addressing these reservations, they were not lifted. However, during 2015, the Commission continued to follow closely the problems with Greece’s compilation of national accounts and kept the technical assistance programme that had been set up in 2014 with the objective of creating an independent and stronger national statistical authority. This work improves the reliability of the Greek GNI data. It is important, as the quality of that data can have an impact on all Member States’ individual contributions in terms of GNI-based own resources (see footnote 4). The number of VAT-based own resources reservations has decreased At the end of the year, there were a total of 85 reservations, compared to 101 in 2014 (see Figure 4.2), of which 69 had been set by the Commission and 16 by the Member States. Of the 69 reservations placed by the Commission, 15 related to infringements of the VAT directive (such as granting exemptions to products or activities that should not be exempted). The number of long-outstanding reservations set by the Commission decreased from ten to six. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The management of traditional own resources by Member States can still be improved Post-clearance audits are an important tool to safeguard the financial interests of the EU,… Customs authorities may check the commercial documents and data related to goods after they have been imported and released for free circulation in the EU. These checks, which include post-clearance audits, are referred to as post-clearance controls. They should be based on risk analyses using criteria developed at the national, EU and, where available, international level. The Commission will continue to work with Member States in line with the Commission Strategy and Action Plan for customs risk management to ensure controls (including post-clearance controls) are effective and risk-based, notably through the further development of EU common risk criteria and standards for financial risks. … but lack of harmonization persists, affecting the results As reported in previous years’ annual reports, we found weaknesses in the identification, selection and inspection of the importers audited by Member States. As an example, in Italy, the selection of post-clearance audits is typically based on the characteristics of individual transactions and not, as recommended in the Commission’s customs audit guide, on the profile of companies. The Commission, in its inspections, regularly examines the post-clearance controls including audits carried out by the Member States and encourages them to make use of the Customs Audit Guide in the conduct of those controls/audits. It will follow up the findings in its regular follow-up of the findings resulting from the Court’s audits and its own inspections. See Commission reply to paragraph 4.14. The quality, scope and results of the post-clearance audits varied substantially across the three Member States visited. A specific case is France, where, for post-clearance audits, a debt can be notified more than three years after it was incurred. This approach differs from that in other Member States because the audits are, according to French legislation, always linked to the prevention and pursuit of customs law infringements and are thus liable to give rise to criminal proceedings. Therefore, the French authorities interrupt the three year time-barring period during post-clearance audits. While there may be acceptable grounds for difference in the scope and results of post-clearance audits carried out across the Member States, the Commission in its inspections in Member States will continue to examine the points raised by the Court concerning the post-clearance audits carried out by the national authorities. The Member States have access to the Customs Audit Guide, the use of which is not mandatory but which should guide them in the conduct of the post-clearance controls they carry out. In accordance with EU customs legislation, a three-year time limit applies for the notification of a customs debt except where the case may give rise to criminal proceedings. The determination of whether a case may give rise to such proceedings differs across the Member States. The proposed Directive on the Union legal framework for customs infringements and sanctions (COM(2013) 884 final), once adopted and applied, is expected to address the issue of systematic use of Article 221(4) of the Community Customs Code and now Article 103(2) Union Customs Code by some Member States. The management of A and B accounts still needs to be improved in some Member States Each Member State sends the Commission a statement of its customs duties and sugar levies established on a monthly basis (the A accounts) and a quarterly statement of established duties which are not included in the A accounts (the B accounts). The Commission detected shortcomings in the Member States’ management of the B accounts in 17 out of 22 Member States visited. As in previous years, we also found inefficiencies in the management of the B accounts. We noted good practices in Lithuania, even though debt recovery was not always successful (see Figure 4.3). In Italy some systematic weaknesses need to be addressed. The Italian authorities could not confirm that all the amounts collected from the B accounts were included in the correct A account statement. The Commission identified two other Member States (Belgium and France) with similar inconsistencies between the A and B accounts statements. These shortcomings did not affect the reliability of the overall amounts as disclosed in the consolidated accounts of the European Union. The Commission examines systematically the management of the B account in each of its inspections in the Member States, and when it finds shortcomings, requests the Member States to rectify them. In the Advisory Committee on the Communities’ Own Resources (ACOR) of December 2015, the Commission services raised again the problem concerning the consistency between the A and B accounts and requested the Member States whose A and В account statements continued to show inconsistencies to supply a plan to resolve them containing a timeline within which they will be resolved and to carry out routine checks to prevent a recurrence of these inconsistencies. For debtor companies registered outside the EU or debts of citizens of countries outside the EU, there is no effective instrument for debt recovery. In six out of 24 cases sampled in Lithuania, the debtors were from non EU countries: citizens of Russia, Belarus or Ukraine and companies registered in Turkey and the British Virgin Islands. Even though the Lithuanian customs authorities carried out all possible recovery procedures within the borders of the EU, in the cases analysed, the debt recovery was not successful. Moreover, in four of the 15 write-off cases that we selected for analysis across all Member States, the authorities had not been able to enforce debt collection due to the fact that the debtors were located in Switzerland, Ukraine or Russia. The Commission considers that the entry into force of the UCC on 1 May 2016 has provided additional instruments to ensure debt collection in the situations examined by the Court. In particular, it expands the use of mandatory guarantees and it provides for optional guarantees and the appointment of jointly liable customs representatives based in the EU. The number of open points remains significant The Commission’s inspections in Member States may result in the national authorities having to take steps because of non-compliance with EU customs rules. These are known as open points. These open points can have a potential financial impact. At the end of 2015 a total of 325 points (see Figure 4.2) were open. The Commission services follow up open points in a timely and prompt manner. They will only close a point when they are satisfied that adequate measures to resolve it have been taken and when any traditional own resources due have been made available, with interest where applicable. However, complex points (e.g. requiring a change of national legislation or instructions, organisational or IT changes in national services or infringement proceedings) take time to resolve. Of the 325 points shown by the Court as open at the end of 2015, only 199 have a potential financial impact. EEA/EFTA contributions and correction mechanisms calculations contained minor errors We identified minor errors in two areas: 4.20.

(a) the 2015 Commission’s calculation of the EEA/EFTA contributions, and

(b) the 2015 Commission’s calculation of correction mechanisms.

(b) The Commission has taken measures to improve its checks and to avoid errors identified by the Court. For example, new checklists were introduced, procedures for preparation of inputs improved and the double checking of documents further strengthened, especially with regards to inputs used. Moreover, a more detailed description of the calculations providing better guidance for new staff involved was developed.

The errors were the result of incorrect calculation formula and inputs from outdated databases. These errors are not material, but they indicate that the current systems for performing and reviewing these calculations should be improved. Annual activity reports provide a fair assessment The 2015 annual activity reports by DG Budget and Eurostat provide a fair assessment of financial management in relation to the legality and regularity of underlying transactions concerning own resources and other revenue. The information provided corroborates our observations and conclusions. CONCLUSION AND RECOMMENDATIONS The conclusion for 2015 Overall audit evidence indicates that revenue is not affected by a material level of error. In particular for revenue:

we assessed the examined systems as effective for GNI and VAT-based own resources;

we assessed the examined systems as overall effective for TOR. The key internal controls in Member States visited are assessed as partially effective;

we found no errors in the transactions we tested.

Recommendations Annex 4.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, we presented seven recommendations. Of these recommendations, four were implemented in most respects and three were implemented in some respects. Following this review and the findings and conclusions for 2015, we recommend that the Commission: Traditional own resources

Recommendation 1: take the necessary steps to ensure that economic operators receive a similar treatment in all Member States as regards the time-barring of the debt notifications following a post-clearance audit;

The Commission partially accepts the recommendation. The Commission is of the view that its proposal for a Directive on the Union legal framework for customs infringements and sanctions (COM(2013) 884 final) addresses the observations of the Court and that the situation is expected to improve once the proposal is adopted and applied. However, the outcome depends on the decision of the legislative authority.

Recommendation 2: provide Member States with guidance to improve their management of the items recorded in the B accounts;

The Commission accepts the recommendation. Its services, in their inspections in the Member States, regularly provide guidance to the national authorities concerning the management of the B account and promote best practice.

Recommendation 3: ensure that Member States correctly declare and make available the amounts collected from the B accounts;

The Commission accepts the recommendation. It will continue to ensure that the Member States correctly declare and make available the amounts recovered from the B accounts.

Recommendation 4: facilitate to the extent possible the recovery of customs debts by the Member States, where the debtors are not based in an EU Member State.

The Commission accepts the recommendation. The Commission will in due course examine and assess the effectiveness of the new UCC in addressing the issues identified by the Court. See also the Commission reply to Figure 4.3. EEA/EFTA contributions and correction mechanisms

Recommendation 5: improve the checks on the calculations of the EEA/EFTA contributions and correction mechanisms.

The Commission accepts the recommendation. In order to improve the checks, it has already improved and strengthened the procedures and checklists.

ANNEX

Anx. 41. 2015 2014 SIZE AND STRUCTURE OF THE SAMPLE Total transactions: 55 55 ESTIMATED IMPACT OF QUANTIFIABLE ERRORS Estimated level of error 0,0 % 0,0 % Upper Error Limit (UEL) 0,0 % Lower Error Limit (LEL) 0,0 %

ANNEX 4.1 RESULTS OF TRANSACTION TESTING FOR REVENUE

2015 2014 SIZE AND STRUCTURE OF THE SAMPLE Total transactions: 55 55 ESTIMATED IMPACT OF QUANTIFIABLE ERRORS Estimated level of error 0,0 % 0,0 % Upper Error Limit (UEL) 0,0 % Lower Error Limit (LEL) 0,0 %

Anx. 42. Year Court recommendation Court's analysis of the progress made Commission reply Fully implemented Being implemented Not implemented Not applicable Insufficient evidence In most respects In some respects 2013 The Commission should encourage Member States to provide more clarification on the methodologies they use for the compilation of data in the area of the non-observed economy and promote harmonisation between Member States in this area. X The Commission should put in place and closely monitor a detailed action plan with clear milestones to address the problems in the compilation of Greece’s national accounts. X The existing statistical action plan for Greece, JOSGAP, will be supplemented by a specific plan for National Accounts during 2016. The Commission should establish minimum risk analysis standards for the customs post-clearance audits, including building upon the information in the existing database of imports, in order to allow Member States to better target risky importers. X The Commission accepted the recommendation. The new version of the Customs Audit Guide (2014) sets out risk indicators for the post-clearance audit. Changes to the existing database of imports are planned but will not be fully operational until 2018. The upgrade of the existing Surveillance 2 system to Surveillance 3 is subject of the implementation of the UCC (UCC-Delegated Act and Implementing Act) and is being carried out taking into account the Work Programme for the UCC (COM Implementing Decision 2014/255/EU). The Commission should encourage Member States to use the existing guidance, and monitor, the implementation of Member States’ post-clearance audits. X The Commission accepted the recommendation. Under the current legal framework the carrying out of controls is a Member State competence. In addition, the Commission will, in the course of its inspections of the customs controls carried out, encourage the Member States to use the existing guidance and, where weaknesses are found in the conduct of post-clearance audits, request the Member States to take remedial measures. The Commission recalls that the revised Customs Audit Guide was approved in March 2014. The Commission considers that it has been fully implementing this recommendation. However, the recommendation also refers to a continuous process that the Commission keeps on pursuing. It will continue to encourage Member States to use the existing guidance. 2013 and 2012 The Commission should encourage Member States to correctly use A and B accounts and to ensure that they are demonstrably complete and correct. X The Commission has done and will continue to do so. It will continue to examine the use of the A and B accounts in the course of its inspections and will request the Member States to ensure that they are complete and correct. 2012 The Commission should review its control framework for the verification of GNI data including carrying out a structured and formalised cost-benefit analysis, carrying out in-depth verification of material and risky GNI components, limiting the use of general reservations and setting materiality criteria for placing reservations. X In April 2016, the Commission published its plans for the forthcoming verification cycle. These include a detailed risk assessment, a cost-benefit analysis and many other issues raised by the Court in recent years. Improvements concerning reservation management and materiality have already been implemented. The Commission should encourage Member States to strengthen customs supervision in order to maximise the amount of TOR collected. X The Commission has done and will continue to do so. In the course of its regular TOR inspections, it will continue to verify that the Member States have put in place appropriate control frameworks to protect the financial interests of the EU in the area of traditional own resources.

ANNEX 4.2 FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR REVENUE

Year Court recommendation Court's analysis of the progress made Commission reply Fully implemented Being implemented Not implemented Not applicable Insufficient evidence In most respects In some respects 2013 The Commission should encourage Member States to provide more clarification on the methodologies they use for the compilation of data in the area of the non-observed economy and promote harmonisation between Member States in this area. X The Commission should put in place and closely monitor a detailed action plan with clear milestones to address the problems in the compilation of Greece’s national accounts. X The existing statistical action plan for Greece, JOSGAP, will be supplemented by a specific plan for National Accounts during 2016. The Commission should establish minimum risk analysis standards for the customs post-clearance audits, including building upon the information in the existing database of imports, in order to allow Member States to better target risky importers. X The Commission accepted the recommendation. The new version of the Customs Audit Guide (2014) sets out risk indicators for the post-clearance audit. Changes to the existing database of imports are planned but will not be fully operational until 2018. The upgrade of the existing Surveillance 2 system to Surveillance 3 is subject of the implementation of the UCC (UCC-Delegated Act and Implementing Act) and is being carried out taking into account the Work Programme for the UCC (COM Implementing Decision 2014/255/EU). The Commission should encourage Member States to use the existing guidance, and monitor, the implementation of Member States’ post-clearance audits. X The Commission accepted the recommendation. Under the current legal framework the carrying out of controls is a Member State competence. In addition, the Commission will, in the course of its inspections of the customs controls carried out, encourage the Member States to use the existing guidance and, where weaknesses are found in the conduct of post-clearance audits, request the Member States to take remedial measures. The Commission recalls that the revised Customs Audit Guide was approved in March 2014. The Commission considers that it has been fully implementing this recommendation. However, the recommendation also refers to a continuous process that the Commission keeps on pursuing. It will continue to encourage Member States to use the existing guidance. 2013 and 2012 The Commission should encourage Member States to correctly use A and B accounts and to ensure that they are demonstrably complete and correct. X The Commission has done and will continue to do so. It will continue to examine the use of the A and B accounts in the course of its inspections and will request the Member States to ensure that they are complete and correct. 2012 The Commission should review its control framework for the verification of GNI data including carrying out a structured and formalised cost-benefit analysis, carrying out in-depth verification of material and risky GNI components, limiting the use of general reservations and setting materiality criteria for placing reservations. X In April 2016, the Commission published its plans for the forthcoming verification cycle. These include a detailed risk assessment, a cost-benefit analysis and many other issues raised by the Court in recent years. Improvements concerning reservation management and materiality have already been implemented. The Commission should encourage Member States to strengthen customs supervision in order to maximise the amount of TOR collected. X The Commission has done and will continue to do so. In the course of its regular TOR inspections, it will continue to verify that the Member States have put in place appropriate control frameworks to protect the financial interests of the EU in the area of traditional own resources.

Anx. 43. The GNI process specific and general reservations are not included in the table. Source: European Court of Auditors. Member State GNI reservations (situation at 31.12.2015) VAT reservations (situation at 31.12.2015) TOR open points (situation at 31.12.2015) Belgium 1 4 22 Bulgaria 1 2 4 Czech Republic 0 0 3 Denmark 0 3 16 Germany 1 4 8 Estonia 0 1 3 Ireland 1 6 8 Greece 12 7 28 Spain 0 2 18 France 1 3 38 Croatia 0 0 3 Italy 0 4 12 Cyprus 0 0 6 Latvia 12 1 2 Lithuania 0 0 3 Luxembourg 0 3 1 Hungary 1 1 9 Malta 1 0 4 Netherlands 0 11 48 Austria 0 4 5 Poland 11 4 7 Portugal 0 2 13 Romania 12 3 16 Slovenia 0 0 4 Slovakia 0 0 3 Finland 0 7 10 Sweden 0 3 8 United Kingdom 1 10 23 TOTAL 31.12.2015 55 85 325 TOTAL 31.12.2014 239 101 348

ANNEX 4.3 NUMBER OF OUTSTANDING GNI RESERVATIONS, VAT RESERVATIONS AND TOR OPEN POINTS BY MEMBER STATE AT 31.12.2015

The GNI process specific and general reservations are not included in the table. Source: European Court of Auditors. Member State GNI reservations (situation at 31.12.2015) VAT reservations (situation at 31.12.2015) TOR open points (situation at 31.12.2015) Belgium 1 4 22 Bulgaria 1 2 4 Czech Republic 0 0 3 Denmark 0 3 16 Germany 1 4 8 Estonia 0 1 3 Ireland 1 6 8 Greece 12 7 28 Spain 0 2 18 France 1 3 38 Croatia 0 0 3 Italy 0 4 12 Cyprus 0 0 6 Latvia 12 1 2 Lithuania 0 0 3 Luxembourg 0 3 1 Hungary 1 1 9 Malta 1 0 4 Netherlands 0 11 48 Austria 0 4 5 Poland 11 4 7 Portugal 0 2 13 Romania 12 3 16 Slovenia 0 0 4 Slovakia 0 0 3 Finland 0 7 10 Sweden 0 3 8 United Kingdom 1 10 23 TOTAL 31.12.2015 55 85 325 TOTAL 31.12.2014 239 101 348

CHAPTER 5 Competitiveness for growth and jobs

Txt. 1. THE COURT’S OBSERVATIONS INTRODUCTION This chapter presents our findings for Competitiveness for growth and jobs. Key information on the activities covered and the spending in 2015 is provided in Figure 5.1. Brief description of Competitiveness for growth and jobs The objectives of the spending include improving research and innovation, enhancing education systems and promoting employment, ensuring a digital single market, promoting renewable energy and energy efficiency, modernising the transport sector and improving the business environment, especially for small and medium-sized enterprises (SMEs). Research and innovation accounts for 62 % of the spending, through the Seventh Framework Programme for Research and Technological Development 2007-2013 (the Seventh Research Framework Programme) and Horizon 2020 — the Framework Programme for Research and Innovation 2014-2020 (Horizon 2020). Other major spending instruments are: the Lifelong Learning Programme and Erasmus+ in the fields of education, training, youth and sport; the Galileo, EGNOS and Copernicus Space programmes; the Trans-European Networks-Transport programme, which provides finance for the development of transport infrastructure; the European Energy Programme for Recovery in support of projects in the energy sector; and the Connecting Europe Facility, which supports the development of networks in the sectors of transport, telecommunications and energy. Almost 90 % of the spending takes the form of grants to private or public beneficiaries participating in projects. The Commission makes advances upon signature of a grant agreement or financing decision. The Commission reimburses the EU-funded costs reported by beneficiaries, deducting advances paid. The principal risk to the regularity of transactions is that beneficiaries declare ineligible costs, which are neither detected nor corrected before reimbursement by the Commission. MFF sub-heading 1.a — Competitiveness for growth and jobs — Key information 2015 (billion euro) Source: 2015 consolidated accounts of the European Union. Total payments for the year 16,8 - advances 9,5 + clearings of advances 7,2 Audited population, total 14,5 THE COURT’S OBSERVATIONS Audit scope and approach Annex 1.1 sets out the key elements of our audit approach and methodology. For the audit of Competitiveness for growth and jobs, the following specific points should be noted: (a) We examined a sample of 150 transactions as defined in Annex 1.1, paragraph 7. The sample is designed to be representative of the range of spending under the MFF sub-heading. It consisted of: 88 transactions for research and innovation; 18 transactions for education, training, youth and sport and 44 transactions for other programmes and activities. Txt. 2. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES (b) For new programmes, such as Horizon 2020 and Erasmus+, almost all payments in 2015 were advances, which are excluded from our sampling. Our transaction testing largely focused on payments to projects under the 2007-2013 programmes, especially the Seventh Research Framework Programme and the Lifelong Learning Programme. (c) We assessed the annual activity reports of the Directorate-General for Research and Innovation (DG RTD), the Directorate-General for Internal Market, Industry, Entrepreneurship and SMEs (DG GROW) and the Directorate-General for Education and Culture (DG EAC). (d) We examined the establishment of the Commission’s Common Support Centre for the management of research and innovation spending and the implementation of the Commission’s ex-post audit strategies for the Seventh Research Framework Programme and Horizon 2020. Txt. 3. REGULARITY OF TRANSACTIONS Annex 5.1 contains a summary of the results of transaction testing. Of the 150 transactions examined, 72 (48 %) were affected by error. On the basis of the 38 errors we have quantified, the estimated level of error is 4,4 %. The error rate reported by the Court is one indicator of the effectiveness of the implementation of EU expenditure. However, the Commission has a multiannual control strategy. On this basis its services estimate a residual error rate, which takes account of recoveries, corrections and the effects of all their controls and audits over the period of implementation of the programme. Figure 5.2 presents the extent to which different types of errors contributed to our estimated level of error for 2015. In research and innovation projects, personnel costs and indirect costs account for most of the project costs. In order to accommodate the diversity of cost structures in the European research landscape, a conscious policy choice has been made to reimburse these cost categories on the basis of real costs, rather than unit costs or flat rates. That explains why they are prone to error, which is reflected in the graph. Horizon 2020 has been conceived to address, to the extent possible, these sources of error through the introduction of a number of simplifications — see paragraph 5.12. Competitiveness for growth and jobs — Breakdown of the estimated level of error Source: European Court of Auditors. THE COURT’S OBSERVATIONS For 10 out of 38 transactions subject to quantified error, we reported errors exceeding 20 % of the examined items. These 10 cases (9 from the Seventh Research Framework Programme and one from the 2007-2013 Competitiveness and Innovation Programme) account for 77 % of the overall estimated level of error for Competitiveness for growth and jobs in 2015. For research and innovation spending, the type and range of errors that we found are similar to those which we have detected throughout the course of the Seventh Research Framework Programme. We found fewer quantified errors in payments for other programmes and activities compared to previous years. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES In 16 cases of quantifiable errors, the Commission, national authorities or independent auditors had sufficient information to prevent, or detect and correct the errors before accepting the expenditure. If all this information had been used to prevent, or detect and correct errors, the estimated level of error for this chapter would have been 0,6 percentage points lower. The Commission has a sound system of ex ante controls in place including detailed automated checklists, written guidance and continuous training. The improvement of this system without imposing additional administrative burdens on beneficiaries, and whilst ensuring that payments to researchers are made promptly, is a constant challenge. The Court’s findings have been and will be used to make further improvements to ex ante controls. As regards independent auditors certifying cost claims, which account for 15 of the 16 cases mentioned by the Court, this is a well-known issue, addressed in previous reports. In order to follow up on the Court’s recommendations, the Commission has organised a series of meetings targeting beneficiaries and independent certifying auditors (> 300) to raise awareness of the most common errors. In addition, feedback has been provided to certifying auditors who have made errors, and a more didactic template for audit certificates has been provided in Horizon 2020. For research, audit certificates are estimated to reduce the error rate by 50 % compared to uncertified claims. So while it is recognised that they do not identify every error, they are an important tool to reduce the overall error rate. Most errors concern overcharged personnel and indirect costs Most of the quantified errors which we found (33 out of 38) concerned the reimbursement of ineligible personnel and indirect costs declared by beneficiaries. We detected these errors mainly, but not exclusively, in projects for the Seventh Research Framework Programme (see Figure 5.3). The Commission’s own audits have also shown that the type and level of errors remain stable as a result of the complex FP7 rules. However, as all FP7 contracts have been signed, the further modification of the legal framework is no longer an option. Over the course of FP7, however, the Commission has attempted to simplify the system within the existing legal framework, for example the simplification measures adopted by the Commission on 24 January 2011 (Decision C(2011) 174). Horizon 2020 includes a radical simplification of the legal framework, in order to meet the expectations of both stakeholders and legislative authorities, focusing in particular on personnel and indirect costs as the main sources of error. Example of the reimbursement of ineligible personnel costs We examined a cost statement of 250000 euro from a beneficiary working with eight partners on a project for the development of cloud computing services under the Information and Communication Technologies Policy Support Programme within the 2007-2013 Competitiveness and Innovation Programme. We found that the personnel costs charged to the project were overstated as they were calculated using a standard number of working hours which was higher than the actual number of hours. The salary cost also included ineligible bonus payments. Example of the reimbursement of ineligible personnel costs The Commission will recover the unduly paid amounts. The overstating of hourly rates was caused by errors in the calculation of productive hours. The rules for this calculation have been simplified in Horizon 2020 to avoid this sort of error. Almost all of the errors we found in cost statements were due to beneficiaries misinterpreting the complex eligibility rules or incorrectly calculating their eligible costs. We identified two cases of suspected fraud, where the beneficiaries appeared to have intentionally overstated their eligible project costs. In view of the complexity of the rules, the Commission has introduced many simplifications into Horizon 2020. These include: (5) one funding rate per action, identified in the call for proposals, (6) a flat rate for indirect costs, (7) simplified time recording and greater use of beneficiaries’ practices, (8) more flexible rules for third parties and subcontracting, (9) simpler and clearer rules for calculation of productive hours (and so personnel costs) Txt. 4. Where policy objectives allow for even simpler rules and mechanisms, these have been introduced. European Research Council and Marie Skłodowska Curie grants are straightforward — they largely concern public bodies, use flat rates and lump sums more widely, and the majority are single beneficiary grants. This limits the scope for misinterpretation, and error rates are below 2 % for these schemes. For personnel costs, beneficiaries often incorrectly calculated the hourly rates of their staff charged to the project or were unable to provide adequate evidence of the actual time spent by their staff on the project. In five cases, beneficiaries erroneously declared the use of external consultants as personnel costs (see Figure 5.4). The rules for the use of consultants has proven to be difficult in FP7, the rules have been simplified in Horizon 2020. Example of the reimbursement of ineligible personnel costs We examined a cost statement of 85000 euro from a beneficiary working with 11 partners on a project under the Seventh Research Framework Programme for the development of environmental monitoring systems. The beneficiary hired consultants to implement the project. We found that the consultants did not meet the criteria to be considered as in-house consultants, especially as the cost of employing the consultants was significantly higher than the personnel costs of employees of the same category at the beneficiary. In addition, the personnel costs claimed for one staff member were not justified by any supporting evidence. Example of the reimbursement of ineligible personnel costs The rules for consultants have been simplified in Horizon 2020 to avoid or reduce this sort of error. For indirect costs, beneficiaries committing errors in declaring actual indirect costs used an incorrect methodology or included ineligible costs in the calculation of the amount of their overheads attributable to the project (see Figure 5.5). For beneficiaries declaring a flat rate of indirect costs, based on a percentage of direct costs, any error in their direct costs led to a proportionate error in their eligible indirect costs. The calculation of actual indirect costs is complex and a regular source of errors. The rules of Horizon 2020 have been simplified and a flat rate of overheads will apply in all cases, so avoiding these errors. Example of the reimbursement of ineligible indirect costs We examined a cost statement of 146000 euro from a beneficiary working with eleven partners on a project under the Seventh Research Framework Programme for the development of enhanced geographical management information systems. We found that part of the personnel costs declared by the beneficiary had actually been deployed on a different project. The beneficiary included ineligible items in its calculation of indirect costs (personnel costs not linked to administrative support, marketing costs, and office and travel costs which were not linked to the research activities). Example of the reimbursement of ineligible indirect costs Regarding indirect costs, the example shows the complexity of calculating actual indirect costs. This is why Horizon 2020 introduces a flat rate of overheads for all projects. The Horizon 2020 rules were simplified. However, in our 2014 annual report we pointed to increased risks of irregular payments in cases where researchers receive additional remuneration; or where participants make use of a large research infrastructure. We recommended that the Commission develop an appropriate control strategy, including adequate checks of costs declared under those specific eligibility criteria. In the limited testing we performed for Horizon 2020 in 2015, we noted that the risk related to additional remuneration occurred in practice. The Commission has set up an audit strategy for Horizon 2020 and is also developing control strategies following the Court’s recommendations. Regarding additional remuneration in non-profit beneficiaries, the Commission developed rules that it hoped would find a balance between simplification, certainty, and reasonable assistance to researchers while ensuring sound financial management. The Court's observation aligns with the Commission’s own experience, and solutions are actively being sought with the organisations and Member States involved. Breaches of procurement rules Although we found 6 cases of error in the 10 procurements we examined, compliance with procurement rules improved significantly compared to the previous year. In one case, we identified a serious failure to comply with the rules on public procurement, and therefore the declared costs for the audited contract were ineligible. The remaining errors included cases of non-compliance with information and publication requirements, incorrect application of the selection criteria and shortcomings in the tender specifications. These procedural errors are not quantified and do not contribute to our estimated error rate. EXAMINATION OF ANNUAL ACTIVITY REPORTS The Commission’s annual activity reports corroborate our findings and conclusions We examined the annual activity reports for DG RTD, DG GROW and DG EAC. We consider that the reports provide a fair assessment of financial management in relation to the regularity of transactions, and the information provided corroborates our findings and conclusions in most respects. For example, the annual activity report of DG RTD includes a reservation on payments in reimbursement of cost claims under the Seventh Research Framework Programme. DG RTD and DG GROW issued new reservations for payments where they estimated that the residual error rate (the level of undetected and uncorrected errors remaining at the end of a programme) will exceed 2 %. DG RTD introduced a reservation concerning payments in reimbursement of cost claims for the Research Fund for Coal and Steel, as the residual error rate was estimated at around 3 %. DG GROW introduced a reservation concerning payments in reimbursement of cost claims for the Competiveness and Innovation 2007-2013 Framework Programme, where the estimated residual error rate was 6,1 %. For both the Research Fund for Coal and Steel and the Competitiveness and Innovation 2007-2013 Framework Programme, the main reason for errors is the complexity of the rules and the fact that because the eligibility conditions are similar but not exactly the same as those for the Seventh Research Framework Programme, beneficiaries make errors by applying the rules for the Seventh Research Framework Programme in the calculation of their cost statements. The rules for the Research Fund for Coal and Steel, a small independent scheme within the Directorate-General for Research and Innovation (expenditure of 44 million euro in 2015) are being harmonised with those of the main Framework Programme wherever possible. EXAMINATION OF SELECTED MANAGEMENT AND CONTROL SYSTEMS Common Support Centre for research and innovation spending Responsibility for research and innovation spending is shared by numerous implementing bodies: eight Commission directorates-general, seven joint undertakings, four executive agencies, one decentralised agency, and several international organisations working under delegation agreements. With a view to efficient and harmonised programme implementation, and taking into account our recommendation, in January 2014 the Commission established a Common Support Centre for the management of Horizon 2020. The Centre is hosted by DG RTD and provides common services in legal support, ex-post audits (through the Common Audit Service), IT systems and operations, business processes and programme information and data. The Common Support Centre is a positive step towards streamlining IT systems and operations, harmonising grant management and ensuring that the various implementing bodies apply Horizon 2020 rules consistently. However, the European Institute of Innovation and Technology, which is expected to manage around 2,3 billion euro or 3 % of the Horizon 2020 budget, is currently not participating in this management and control framework. Common Audit Service For research and innovation spending, the Commission has recognised the need to reduce administrative burdens on beneficiaries while maintaining an appropriate balance between trust and control. Therefore, the Commission has reduced checks before payments and seeks to obtain most of its assurance through its ex-post audits of reimbursed costs. The ex-post audits are conducted either by the Commission’s own auditors from the Common Audit Service or by independent auditors working on behalf of the Commission under a framework contract. Although the Common Audit Service was set up as part of the Common Support Centre in January 2014, delays were experienced in recruiting staff and most of the newly recruited staff only joined the Service towards the end of 2015. Nevertheless, the objectives for the number of audits launched and the number of audits focused on detecting fraud were met. Implementation of the objective concerning the number of finalised audits was slightly below target, with 456 audits closed against a target of 473 audits (96 %). The different implementing bodies have had difficulties in finding a common position on how to implement some of the audit recommendations for recovering ineligible costs for projects under the Seventh Research Framework Programme. This increases the risk of inconsistent treatment of beneficiaries and raises legal uncertainty. For Horizon 2020, discussions are ongoing to establish a process to ensure consistency for recoveries while respecting the obligations of Authorising Officers. The different services of the research family have also agreed on a coordination of the extrapolation process to improve harmonisation for the Seventh Framework Programme. Implementation of ex-post audit results for the Seventh Research Framework Programme By the end of 2015, the Commission had audited reimbursed costs amounting to 2,5 billion euro or 8 % of payments made of 30,9 billion euro for the Seventh Research Framework Programme. The Commission had closed 3334 audits out of the 4116 planned in its audit strategy which continues until the end of 2016. The Commission has calculated a statistically valid error rate of 4,5 % for the Seventh Research Framework Programme. At the end of 2015, the cumulative adjustments following the Commission’s ex-post audits amounted to 107 million euro, out of which 79,4 million euro was recovered (74 %) (see paragraphs 1.40 to 1.43). In cases where systematic errors are identified, the Commission applies extrapolated corrections. At the end of 2015, the Commission had implemented 5868 extrapolations out of 8592 cases, an implementation rate of 68 % (compared to 74,9 % at the end of 2014). The rate of extrapolation is considered to be satisfactory at this stage of the Seventh Framework Programme. The rate fell at the end of 2015 as many audits were closed at this time, but the extrapolations will take place in 2016. Taking into account errors corrected as a result of the extrapolation of audit results, the Commission has calculated a residual error rate of 2,8 % for the Seventh Research Framework Programme. This represents the expected level of error remaining undetected and uncorrected after all payments for the programme have been completed. Taking into consideration the multiannual nature of the programme, we consider that the Commission’s ex-post audit strategy for the Seventh Research Framework Programme provides an appropriate basis for calculating the residual error rate for spending under the programme. The number of closed audits at the end of 2015 was in line with expectations. Horizon 2020 ex-post audit strategy For Horizon 2020, the Commission is required to implement an audit strategy based on the financial audit of a representative sample of spending across the programme as a whole, supplemented by a selection based on an assessment of the risks related to spending. The main objective is to provide assurance by assessing the legality and regularity of payments in reimbursement of cost claims and by attaining a reasonable estimate of the residual error rate at the end of the programme. The Commission approved its Horizon 2020 ex-post audit strategy in December 2015, including a total of 4400 audits to be performed. The Commission has set out its intention that a maximum of seven percent of the number of beneficiaries will be audited under the strategy, to reduce the audit burden on the latter. Despite some delays in launching the framework contract for outsourcing Horizon 2020 audits, the implementation of the strategy is expected to begin as scheduled in 2016. CONCLUSION AND RECOMMENDATIONS The conclusion for 2015 Overall audit evidence indicates that spending on Competitiveness for growth and jobs is affected by a material level of error For this MFF sub-heading area, testing of transactions indicates that the estimated level of error present in the population is 4,4 % (see Annex 5.1). Please see the Commission reply to paragraph 5.6. Recommendations Annex 5.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, we presented eight recommendations. Of these, we consider one as not applicable, as its implementation needs to be assessed over several years. The Commission fully implemented one recommendation, while six were implemented in most respects. Following this review and the findings and conclusions for 2015, we recommend that: (10) Recommendation 1: the Commission, national authorities and independent auditors use all the relevant information available to prevent, or detect and correct errors before reimbursement (see paragraph 5.10); Txt. 5. The Commission accepts the recommendation. It has a sound system of ex ante controls in place including detailed automated checklists, written guidance and continuous training. The improvement of this system without imposing additional administrative burdens on beneficiaries, and whilst ensuring that payments to researchers are made promptly, is a constant challenge. The Court’s findings have been and will be used to make further improvements to ex ante controls. (11) Recommendation 2: the Commission issue guidance to beneficiaries on the specific differences in Horizon 2020 compared to the Seventh Research Framework Programme and similar programmes (see paragraphs 5.15 and 5.20); Txt. 6. The Commission accepts the recommendation. It is already providing guidance on this question in a range of ways. The differences between FP7 and Horizon 2020 rules are specifically highlighted in the annotated model grant agreement. An extensive communication campaign on Horizon 2020 legal and financial rules was launched in 2015, with 16 events in the Member States. These events were attended by 2046 representatives of Horizon 2020 beneficiaries and certifying auditors. Another 14 events are planned for 2016. The Common Support Centre provides guidance on legal and financial issues often pointing to the differences between FP7 and Horizon 2020 rules. (12) Recommendation 3: the Commission issue common guidelines to the implementing bodies for research and innovation spending, in order to ensure consistent treatment of beneficiaries when applying audit recommendations for the recovery of ineligible costs (see paragraph 5.26); Txt. 7. The Commission accepts the recommendation. The different services have agreed on an extrapolation process to improve harmonisation for the Seventh Framework Programme. For Horizon 2020, discussions are ongoing to establish a process to ensure consistency of audit implementation while respecting the obligations of Authorising Officers. (13) Recommendation 4: the Commission closely monitor the implementation of extrapolated corrections based on its ex-post audits of reimbursed costs under the Seventh Research Framework Programme (see paragraph 5.29). Txt. 8. The Commission accepts the recommendation. It closely monitors the implementation of extrapolation cases, and is satisfied with progress at this stage of the Seventh Framework Programme. The implementation rate fell at the end of 2015 as many audits were closed at this time, but the extrapolations will take place in 2016.

THE COURT’S OBSERVATIONS INTRODUCTION This chapter presents our findings for Competitiveness for growth and jobs. Key information on the activities covered and the spending in 2015 is provided in Figure 5.1. Brief description of Competitiveness for growth and jobs The objectives of the spending include improving research and innovation, enhancing education systems and promoting employment, ensuring a digital single market, promoting renewable energy and energy efficiency, modernising the transport sector and improving the business environment, especially for small and medium-sized enterprises (SMEs). Research and innovation accounts for 62 % of the spending, through the Seventh Framework Programme for Research and Technological Development 2007-2013 (the Seventh Research Framework Programme) and Horizon 2020 — the Framework Programme for Research and Innovation 2014-2020 (Horizon 2020). Other major spending instruments are: the Lifelong Learning Programme and Erasmus+ in the fields of education, training, youth and sport; the Galileo, EGNOS and Copernicus Space programmes; the Trans-European Networks-Transport programme, which provides finance for the development of transport infrastructure; the European Energy Programme for Recovery in support of projects in the energy sector; and the Connecting Europe Facility, which supports the development of networks in the sectors of transport, telecommunications and energy. Almost 90 % of the spending takes the form of grants to private or public beneficiaries participating in projects. The Commission makes advances upon signature of a grant agreement or financing decision. The Commission reimburses the EU-funded costs reported by beneficiaries, deducting advances paid. The principal risk to the regularity of transactions is that beneficiaries declare ineligible costs, which are neither detected nor corrected before reimbursement by the Commission. MFF sub-heading 1.a — Competitiveness for growth and jobs — Key information 2015 (billion euro) Source: 2015 consolidated accounts of the European Union. Total payments for the year 16,8 - advances 9,5 + clearings of advances 7,2 Audited population, total 14,5 THE COURT’S OBSERVATIONS Audit scope and approach Annex 1.1 sets out the key elements of our audit approach and methodology. For the audit of Competitiveness for growth and jobs, the following specific points should be noted:

(a) We examined a sample of 150 transactions as defined in Annex 1.1, paragraph 7. The sample is designed to be representative of the range of spending under the MFF sub-heading. It consisted of: 88 transactions for research and innovation; 18 transactions for education, training, youth and sport and 44 transactions for other programmes and activities.

THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES

(b) For new programmes, such as Horizon 2020 and Erasmus+, almost all payments in 2015 were advances, which are excluded from our sampling. Our transaction testing largely focused on payments to projects under the 2007-2013 programmes, especially the Seventh Research Framework Programme and the Lifelong Learning Programme.

(c) We assessed the annual activity reports of the Directorate-General for Research and Innovation (DG RTD), the Directorate-General for Internal Market, Industry, Entrepreneurship and SMEs (DG GROW) and the Directorate-General for Education and Culture (DG EAC).

(d) We examined the establishment of the Commission’s Common Support Centre for the management of research and innovation spending and the implementation of the Commission’s ex-post audit strategies for the Seventh Research Framework Programme and Horizon 2020.

REGULARITY OF TRANSACTIONS Annex 5.1 contains a summary of the results of transaction testing. Of the 150 transactions examined, 72 (48 %) were affected by error. On the basis of the 38 errors we have quantified, the estimated level of error is 4,4 %. The error rate reported by the Court is one indicator of the effectiveness of the implementation of EU expenditure. However, the Commission has a multiannual control strategy. On this basis its services estimate a residual error rate, which takes account of recoveries, corrections and the effects of all their controls and audits over the period of implementation of the programme. Figure 5.2 presents the extent to which different types of errors contributed to our estimated level of error for 2015. In research and innovation projects, personnel costs and indirect costs account for most of the project costs. In order to accommodate the diversity of cost structures in the European research landscape, a conscious policy choice has been made to reimburse these cost categories on the basis of real costs, rather than unit costs or flat rates. That explains why they are prone to error, which is reflected in the graph. Horizon 2020 has been conceived to address, to the extent possible, these sources of error through the introduction of a number of simplifications — see paragraph 5.12. Competitiveness for growth and jobs — Breakdown of the estimated level of error Source: European Court of Auditors. THE COURT’S OBSERVATIONS For 10 out of 38 transactions subject to quantified error, we reported errors exceeding 20 % of the examined items. These 10 cases (9 from the Seventh Research Framework Programme and one from the 2007-2013 Competitiveness and Innovation Programme) account for 77 % of the overall estimated level of error for Competitiveness for growth and jobs in 2015. For research and innovation spending, the type and range of errors that we found are similar to those which we have detected throughout the course of the Seventh Research Framework Programme. We found fewer quantified errors in payments for other programmes and activities compared to previous years. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES In 16 cases of quantifiable errors, the Commission, national authorities or independent auditors had sufficient information to prevent, or detect and correct the errors before accepting the expenditure. If all this information had been used to prevent, or detect and correct errors, the estimated level of error for this chapter would have been 0,6 percentage points lower. The Commission has a sound system of ex ante controls in place including detailed automated checklists, written guidance and continuous training. The improvement of this system without imposing additional administrative burdens on beneficiaries, and whilst ensuring that payments to researchers are made promptly, is a constant challenge. The Court’s findings have been and will be used to make further improvements to ex ante controls. As regards independent auditors certifying cost claims, which account for 15 of the 16 cases mentioned by the Court, this is a well-known issue, addressed in previous reports. In order to follow up on the Court’s recommendations, the Commission has organised a series of meetings targeting beneficiaries and independent certifying auditors (> 300) to raise awareness of the most common errors. In addition, feedback has been provided to certifying auditors who have made errors, and a more didactic template for audit certificates has been provided in Horizon 2020. For research, audit certificates are estimated to reduce the error rate by 50 % compared to uncertified claims. So while it is recognised that they do not identify every error, they are an important tool to reduce the overall error rate. Most errors concern overcharged personnel and indirect costs Most of the quantified errors which we found (33 out of 38) concerned the reimbursement of ineligible personnel and indirect costs declared by beneficiaries. We detected these errors mainly, but not exclusively, in projects for the Seventh Research Framework Programme (see Figure 5.3). The Commission’s own audits have also shown that the type and level of errors remain stable as a result of the complex FP7 rules. However, as all FP7 contracts have been signed, the further modification of the legal framework is no longer an option. Over the course of FP7, however, the Commission has attempted to simplify the system within the existing legal framework, for example the simplification measures adopted by the Commission on 24 January 2011 (Decision C(2011) 174). Horizon 2020 includes a radical simplification of the legal framework, in order to meet the expectations of both stakeholders and legislative authorities, focusing in particular on personnel and indirect costs as the main sources of error. Example of the reimbursement of ineligible personnel costs We examined a cost statement of 250000 euro from a beneficiary working with eight partners on a project for the development of cloud computing services under the Information and Communication Technologies Policy Support Programme within the 2007-2013 Competitiveness and Innovation Programme. We found that the personnel costs charged to the project were overstated as they were calculated using a standard number of working hours which was higher than the actual number of hours. The salary cost also included ineligible bonus payments. Example of the reimbursement of ineligible personnel costs The Commission will recover the unduly paid amounts. The overstating of hourly rates was caused by errors in the calculation of productive hours. The rules for this calculation have been simplified in Horizon 2020 to avoid this sort of error. Almost all of the errors we found in cost statements were due to beneficiaries misinterpreting the complex eligibility rules or incorrectly calculating their eligible costs. We identified two cases of suspected fraud, where the beneficiaries appeared to have intentionally overstated their eligible project costs. In view of the complexity of the rules, the Commission has introduced many simplifications into Horizon 2020. These include:

one funding rate per action, identified in the call for proposals,

a flat rate for indirect costs,

simplified time recording and greater use of beneficiaries’ practices,

more flexible rules for third parties and subcontracting,

simpler and clearer rules for calculation of productive hours (and so personnel costs)

Where policy objectives allow for even simpler rules and mechanisms, these have been introduced. European Research Council and Marie Skłodowska Curie grants are straightforward — they largely concern public bodies, use flat rates and lump sums more widely, and the majority are single beneficiary grants. This limits the scope for misinterpretation, and error rates are below 2 % for these schemes. For personnel costs, beneficiaries often incorrectly calculated the hourly rates of their staff charged to the project or were unable to provide adequate evidence of the actual time spent by their staff on the project. In five cases, beneficiaries erroneously declared the use of external consultants as personnel costs (see Figure 5.4). The rules for the use of consultants has proven to be difficult in FP7, the rules have been simplified in Horizon 2020. Example of the reimbursement of ineligible personnel costs We examined a cost statement of 85000 euro from a beneficiary working with 11 partners on a project under the Seventh Research Framework Programme for the development of environmental monitoring systems. The beneficiary hired consultants to implement the project. We found that the consultants did not meet the criteria to be considered as in-house consultants, especially as the cost of employing the consultants was significantly higher than the personnel costs of employees of the same category at the beneficiary. In addition, the personnel costs claimed for one staff member were not justified by any supporting evidence. Example of the reimbursement of ineligible personnel costs The rules for consultants have been simplified in Horizon 2020 to avoid or reduce this sort of error. For indirect costs, beneficiaries committing errors in declaring actual indirect costs used an incorrect methodology or included ineligible costs in the calculation of the amount of their overheads attributable to the project (see Figure 5.5). For beneficiaries declaring a flat rate of indirect costs, based on a percentage of direct costs, any error in their direct costs led to a proportionate error in their eligible indirect costs. The calculation of actual indirect costs is complex and a regular source of errors. The rules of Horizon 2020 have been simplified and a flat rate of overheads will apply in all cases, so avoiding these errors. Example of the reimbursement of ineligible indirect costs We examined a cost statement of 146000 euro from a beneficiary working with eleven partners on a project under the Seventh Research Framework Programme for the development of enhanced geographical management information systems. We found that part of the personnel costs declared by the beneficiary had actually been deployed on a different project. The beneficiary included ineligible items in its calculation of indirect costs (personnel costs not linked to administrative support, marketing costs, and office and travel costs which were not linked to the research activities). Example of the reimbursement of ineligible indirect costs Regarding indirect costs, the example shows the complexity of calculating actual indirect costs. This is why Horizon 2020 introduces a flat rate of overheads for all projects. The Horizon 2020 rules were simplified. However, in our 2014 annual report we pointed to increased risks of irregular payments in cases where researchers receive additional remuneration; or where participants make use of a large research infrastructure. We recommended that the Commission develop an appropriate control strategy, including adequate checks of costs declared under those specific eligibility criteria. In the limited testing we performed for Horizon 2020 in 2015, we noted that the risk related to additional remuneration occurred in practice. The Commission has set up an audit strategy for Horizon 2020 and is also developing control strategies following the Court’s recommendations. Regarding additional remuneration in non-profit beneficiaries, the Commission developed rules that it hoped would find a balance between simplification, certainty, and reasonable assistance to researchers while ensuring sound financial management. The Court's observation aligns with the Commission’s own experience, and solutions are actively being sought with the organisations and Member States involved. Breaches of procurement rules Although we found 6 cases of error in the 10 procurements we examined, compliance with procurement rules improved significantly compared to the previous year. In one case, we identified a serious failure to comply with the rules on public procurement, and therefore the declared costs for the audited contract were ineligible. The remaining errors included cases of non-compliance with information and publication requirements, incorrect application of the selection criteria and shortcomings in the tender specifications. These procedural errors are not quantified and do not contribute to our estimated error rate. EXAMINATION OF ANNUAL ACTIVITY REPORTS The Commission’s annual activity reports corroborate our findings and conclusions We examined the annual activity reports for DG RTD, DG GROW and DG EAC. We consider that the reports provide a fair assessment of financial management in relation to the regularity of transactions, and the information provided corroborates our findings and conclusions in most respects. For example, the annual activity report of DG RTD includes a reservation on payments in reimbursement of cost claims under the Seventh Research Framework Programme. DG RTD and DG GROW issued new reservations for payments where they estimated that the residual error rate (the level of undetected and uncorrected errors remaining at the end of a programme) will exceed 2 %. DG RTD introduced a reservation concerning payments in reimbursement of cost claims for the Research Fund for Coal and Steel, as the residual error rate was estimated at around 3 %. DG GROW introduced a reservation concerning payments in reimbursement of cost claims for the Competiveness and Innovation 2007-2013 Framework Programme, where the estimated residual error rate was 6,1 %. For both the Research Fund for Coal and Steel and the Competitiveness and Innovation 2007-2013 Framework Programme, the main reason for errors is the complexity of the rules and the fact that because the eligibility conditions are similar but not exactly the same as those for the Seventh Research Framework Programme, beneficiaries make errors by applying the rules for the Seventh Research Framework Programme in the calculation of their cost statements. The rules for the Research Fund for Coal and Steel, a small independent scheme within the Directorate-General for Research and Innovation (expenditure of 44 million euro in 2015) are being harmonised with those of the main Framework Programme wherever possible. EXAMINATION OF SELECTED MANAGEMENT AND CONTROL SYSTEMS Common Support Centre for research and innovation spending Responsibility for research and innovation spending is shared by numerous implementing bodies: eight Commission directorates-general, seven joint undertakings, four executive agencies, one decentralised agency, and several international organisations working under delegation agreements. With a view to efficient and harmonised programme implementation, and taking into account our recommendation, in January 2014 the Commission established a Common Support Centre for the management of Horizon 2020. The Centre is hosted by DG RTD and provides common services in legal support, ex-post audits (through the Common Audit Service), IT systems and operations, business processes and programme information and data. The Common Support Centre is a positive step towards streamlining IT systems and operations, harmonising grant management and ensuring that the various implementing bodies apply Horizon 2020 rules consistently. However, the European Institute of Innovation and Technology, which is expected to manage around 2,3 billion euro or 3 % of the Horizon 2020 budget, is currently not participating in this management and control framework. Common Audit Service For research and innovation spending, the Commission has recognised the need to reduce administrative burdens on beneficiaries while maintaining an appropriate balance between trust and control. Therefore, the Commission has reduced checks before payments and seeks to obtain most of its assurance through its ex-post audits of reimbursed costs. The ex-post audits are conducted either by the Commission’s own auditors from the Common Audit Service or by independent auditors working on behalf of the Commission under a framework contract. Although the Common Audit Service was set up as part of the Common Support Centre in January 2014, delays were experienced in recruiting staff and most of the newly recruited staff only joined the Service towards the end of 2015. Nevertheless, the objectives for the number of audits launched and the number of audits focused on detecting fraud were met. Implementation of the objective concerning the number of finalised audits was slightly below target, with 456 audits closed against a target of 473 audits (96 %). The different implementing bodies have had difficulties in finding a common position on how to implement some of the audit recommendations for recovering ineligible costs for projects under the Seventh Research Framework Programme. This increases the risk of inconsistent treatment of beneficiaries and raises legal uncertainty. For Horizon 2020, discussions are ongoing to establish a process to ensure consistency for recoveries while respecting the obligations of Authorising Officers. The different services of the research family have also agreed on a coordination of the extrapolation process to improve harmonisation for the Seventh Framework Programme. Implementation of ex-post audit results for the Seventh Research Framework Programme By the end of 2015, the Commission had audited reimbursed costs amounting to 2,5 billion euro or 8 % of payments made of 30,9 billion euro for the Seventh Research Framework Programme. The Commission had closed 3334 audits out of the 4116 planned in its audit strategy which continues until the end of 2016. The Commission has calculated a statistically valid error rate of 4,5 % for the Seventh Research Framework Programme. At the end of 2015, the cumulative adjustments following the Commission’s ex-post audits amounted to 107 million euro, out of which 79,4 million euro was recovered (74 %) (see paragraphs 1.40 to 1.43). In cases where systematic errors are identified, the Commission applies extrapolated corrections. At the end of 2015, the Commission had implemented 5868 extrapolations out of 8592 cases, an implementation rate of 68 % (compared to 74,9 % at the end of 2014). The rate of extrapolation is considered to be satisfactory at this stage of the Seventh Framework Programme. The rate fell at the end of 2015 as many audits were closed at this time, but the extrapolations will take place in 2016. Taking into account errors corrected as a result of the extrapolation of audit results, the Commission has calculated a residual error rate of 2,8 % for the Seventh Research Framework Programme. This represents the expected level of error remaining undetected and uncorrected after all payments for the programme have been completed. Taking into consideration the multiannual nature of the programme, we consider that the Commission’s ex-post audit strategy for the Seventh Research Framework Programme provides an appropriate basis for calculating the residual error rate for spending under the programme. The number of closed audits at the end of 2015 was in line with expectations. Horizon 2020 ex-post audit strategy For Horizon 2020, the Commission is required to implement an audit strategy based on the financial audit of a representative sample of spending across the programme as a whole, supplemented by a selection based on an assessment of the risks related to spending. The main objective is to provide assurance by assessing the legality and regularity of payments in reimbursement of cost claims and by attaining a reasonable estimate of the residual error rate at the end of the programme. The Commission approved its Horizon 2020 ex-post audit strategy in December 2015, including a total of 4400 audits to be performed. The Commission has set out its intention that a maximum of seven percent of the number of beneficiaries will be audited under the strategy, to reduce the audit burden on the latter. Despite some delays in launching the framework contract for outsourcing Horizon 2020 audits, the implementation of the strategy is expected to begin as scheduled in 2016. CONCLUSION AND RECOMMENDATIONS The conclusion for 2015 Overall audit evidence indicates that spending on Competitiveness for growth and jobs is affected by a material level of error For this MFF sub-heading area, testing of transactions indicates that the estimated level of error present in the population is 4,4 % (see Annex 5.1). Please see the Commission reply to paragraph 5.6. Recommendations Annex 5.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, we presented eight recommendations. Of these, we consider one as not applicable, as its implementation needs to be assessed over several years. The Commission fully implemented one recommendation, while six were implemented in most respects. Following this review and the findings and conclusions for 2015, we recommend that:

Recommendation 1: the Commission, national authorities and independent auditors use all the relevant information available to prevent, or detect and correct errors before reimbursement (see paragraph 5.10);

The Commission accepts the recommendation. It has a sound system of ex ante controls in place including detailed automated checklists, written guidance and continuous training. The improvement of this system without imposing additional administrative burdens on beneficiaries, and whilst ensuring that payments to researchers are made promptly, is a constant challenge. The Court’s findings have been and will be used to make further improvements to ex ante controls.

Recommendation 2: the Commission issue guidance to beneficiaries on the specific differences in Horizon 2020 compared to the Seventh Research Framework Programme and similar programmes (see paragraphs 5.15 and 5.20);

The Commission accepts the recommendation. It is already providing guidance on this question in a range of ways. The differences between FP7 and Horizon 2020 rules are specifically highlighted in the annotated model grant agreement. An extensive communication campaign on Horizon 2020 legal and financial rules was launched in 2015, with 16 events in the Member States. These events were attended by 2046 representatives of Horizon 2020 beneficiaries and certifying auditors. Another 14 events are planned for 2016. The Common Support Centre provides guidance on legal and financial issues often pointing to the differences between FP7 and Horizon 2020 rules.

Recommendation 3: the Commission issue common guidelines to the implementing bodies for research and innovation spending, in order to ensure consistent treatment of beneficiaries when applying audit recommendations for the recovery of ineligible costs (see paragraph 5.26);

The Commission accepts the recommendation. The different services have agreed on an extrapolation process to improve harmonisation for the Seventh Framework Programme. For Horizon 2020, discussions are ongoing to establish a process to ensure consistency of audit implementation while respecting the obligations of Authorising Officers.

Recommendation 4: the Commission closely monitor the implementation of extrapolated corrections based on its ex-post audits of reimbursed costs under the Seventh Research Framework Programme (see paragraph 5.29).

The Commission accepts the recommendation. It closely monitors the implementation of extrapolation cases, and is satisfied with progress at this stage of the Seventh Framework Programme. The implementation rate fell at the end of 2015 as many audits were closed at this time, but the extrapolations will take place in 2016.

ANNEX

Anx. 51. 2015 2014 SIZE AND STRUCTURE OF THE SAMPLE Total transactions: 150 166 ESTIMATED IMPACT OF QUANTIFIABLE ERRORS Estimated level of error 4,4 % 5,6 % Upper Error Limit (UEL) 6,7 % Lower Error Limit (LEL) 2,0 %

ANNEX 5.1 RESULTS OF TRANSACTION TESTING FOR COMPETITIVENESS FOR GROWTH AND JOBS

2015 2014 SIZE AND STRUCTURE OF THE SAMPLE Total transactions: 150 166 ESTIMATED IMPACT OF QUANTIFIABLE ERRORS Estimated level of error 4,4 % 5,6 % Upper Error Limit (UEL) 6,7 % Lower Error Limit (LEL) 2,0 %

(a) the yearly cycle of supervisory visits performed by EAC staff at national authorities and national agencies; (b) the minimum percentages and minimum absolute numbers of checks on beneficiaries (primary checks) to be carried out every year by national agencies (type of funding scheme, complexity of rules, beneficiary structure, etc.); (c) national agencies risk management for performing targeted additional checks on beneficiaries on top of the minimum required checks imposed by DG EAC. Anx. 52. Year Court recommendation Court's analysis of the progress made Commission reply Fully implemented Being implemented Not implemented Not applicable Insufficient evidence In most respects In some respects 2013 The Commission should: Recommendation 1: extend and intensify its communication campaign to raise awareness among beneficiaries and independent auditors about the eligibility rules for research spending under FP7; X The Communication campaign has so far reached over 4500 people. The Commission will continue these efforts in the context of providing guidance for Horizon 2020 (see recommendation 3). In the research area, awareness-raising activities alert Commission staff to the particular risks of different types of beneficiary. This relates both to the payment stage (FP7) and the contracting stage (Horizon 2020). For Horizon 2020, this type of information will be built more systematically into the information systems used by the Commission. It is expected that 83 % of the ex post audits for the period 2012-2016 will be selected using different risk factors. The research anti-fraud strategy is another important component of our risk-driven and risk-based targeted controls. Recommendation 2: across the policy group, make its control activities more risk-driven, focusing checks on high-risk beneficiaries (for example entities with less experience of European funding) and reducing the burden of checks on less risky beneficiaries; X See above (response to Recommendation 1). In the education and culture policy area the Commission applies risk-driven activities in the following areas: 2013 Recommendation 3: for the new 2014-2020 programmes for research and other internal policies, provide timely, consistent and clear guidance to beneficiaries and managing authorities in respect of the revised eligibility and control requirements. X 2012 Recommendation 1: further intensify its efforts to address the errors found in interim and final payments and clearings, in particular by reminding beneficiaries and independent auditors of the eligibility rules and the requirement for beneficiaries to substantiate all declared costs. X See also the Commission reply to recommendation 1 of 2013. Recommendation 2: remind research FP project coordinators of their responsibility to distribute the funds received to the other project partners without undue delay. X Recommendation 3: review the cases of weaknesses in ex ante checks identified by the Court in order to assess if the checks require modification. X Weaknesses in ex ante checks identified by different sources (also by the Court) are being addressed by the Commission on a permanent basis. This process is now embedded in the business processes of Horizon 2020. Recommendation 4: reduce delays in the implementation of ex post audits and increase the implementation rate for extrapolation cases. X The Commission is monitoring the implementation of the Commission ex post audit reports on a permanent basis. Furthermore, the annual activity reports (AAR) of the Commission’s DGs provide information on progress on these aspects. 2012 Recommendation 5: reinforce the supervisory and control systems for CIP ICT-PSP. X The Commission (DG CONNECT) has adopted an audit strategy covering the non-research strand of the DG’s spending aiming at providing assurance to the DG’s Director-General as to the management of the non-research funding. In 2015, 92 new CIP ICT PSP audits were launched. In total, 52 non-research audits were closed covering 23,7 million euro. The progress of the implementation of the strategy is monitored on a monthly basis in the Audit Budget and Control (ABC) meeting chaired by the Director-General and attended by representatives of all Directorates. In 2012, simplification measures were adopted for SME owners. In addition, the CIP model agreement was revised following the entry into force of the new Financial Regulation in 2013 to cover extrapolation and third parties.

ANNEX 5.2 FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR COMPETITIVENESS FOR GROWTH AND JOBS

Year Court recommendation Court's analysis of the progress made Commission reply Fully implemented Being implemented Not implemented Not applicable Insufficient evidence In most respects In some respects 2013 The Commission should: Recommendation 1: extend and intensify its communication campaign to raise awareness among beneficiaries and independent auditors about the eligibility rules for research spending under FP7; X The Communication campaign has so far reached over 4500 people. The Commission will continue these efforts in the context of providing guidance for Horizon 2020 (see recommendation 3). In the research area, awareness-raising activities alert Commission staff to the particular risks of different types of beneficiary. This relates both to the payment stage (FP7) and the contracting stage (Horizon 2020). For Horizon 2020, this type of information will be built more systematically into the information systems used by the Commission. It is expected that 83 % of the ex post audits for the period 2012-2016 will be selected using different risk factors. The research anti-fraud strategy is another important component of our risk-driven and risk-based targeted controls. Recommendation 2: across the policy group, make its control activities more risk-driven, focusing checks on high-risk beneficiaries (for example entities with less experience of European funding) and reducing the burden of checks on less risky beneficiaries; X See above (response to Recommendation 1). In the education and culture policy area the Commission applies risk-driven activities in the following areas: 2013 Recommendation 3: for the new 2014-2020 programmes for research and other internal policies, provide timely, consistent and clear guidance to beneficiaries and managing authorities in respect of the revised eligibility and control requirements. X 2012 Recommendation 1: further intensify its efforts to address the errors found in interim and final payments and clearings, in particular by reminding beneficiaries and independent auditors of the eligibility rules and the requirement for beneficiaries to substantiate all declared costs. X See also the Commission reply to recommendation 1 of 2013. Recommendation 2: remind research FP project coordinators of their responsibility to distribute the funds received to the other project partners without undue delay. X Recommendation 3: review the cases of weaknesses in ex ante checks identified by the Court in order to assess if the checks require modification. X Weaknesses in ex ante checks identified by different sources (also by the Court) are being addressed by the Commission on a permanent basis. This process is now embedded in the business processes of Horizon 2020. Recommendation 4: reduce delays in the implementation of ex post audits and increase the implementation rate for extrapolation cases. X The Commission is monitoring the implementation of the Commission ex post audit reports on a permanent basis. Furthermore, the annual activity reports (AAR) of the Commission’s DGs provide information on progress on these aspects. 2012 Recommendation 5: reinforce the supervisory and control systems for CIP ICT-PSP. X The Commission (DG CONNECT) has adopted an audit strategy covering the non-research strand of the DG’s spending aiming at providing assurance to the DG’s Director-General as to the management of the non-research funding. In 2015, 92 new CIP ICT PSP audits were launched. In total, 52 non-research audits were closed covering 23,7 million euro. The progress of the implementation of the strategy is monitored on a monthly basis in the Audit Budget and Control (ABC) meeting chaired by the Director-General and attended by representatives of all Directorates. In 2012, simplification measures were adopted for SME owners. In addition, the CIP model agreement was revised following the entry into force of the new Financial Regulation in 2013 to cover extrapolation and third parties. (a) the yearly cycle of supervisory visits performed by EAC staff at national authorities and national agencies; (b) the minimum percentages and minimum absolute numbers of checks on beneficiaries (primary checks) to be carried out every year by national agencies (type of funding scheme, complexity of rules, beneficiary structure, etc.); (c) national agencies risk management for performing targeted additional checks on beneficiaries on top of the minimum required checks imposed by DG EAC.

(a) the yearly cycle of supervisory visits performed by EAC staff at national authorities and national agencies;

(b) the minimum percentages and minimum absolute numbers of checks on beneficiaries (primary checks) to be carried out every year by national agencies (type of funding scheme, complexity of rules, beneficiary structure, etc.);

(c) national agencies risk management for performing targeted additional checks on beneficiaries on top of the minimum required checks imposed by DG EAC.

CHAPTER 6 Economic, social and territorial cohesion

Txt. 1. THE COURT’S OBSERVATIONS INTRODUCTION This chapter presents our findings for Economic, social and territorial cohesion (MFF heading 1.b). Key information on the activities covered and their spending in 2015 is provided in Figure 6.1. MFF heading 1.b — Key Information 2015 (billion euro) Source: 2015 consolidated accounts of the European Union. Total payments for the year 51,2 - advances 9,1 + clearings of advances 7,4 + disbursements to final recipients from financial instruments under shared management 2,1 + advances used by final beneficiaries of state aid projects 2,3 Audited population, total 53,9 THE COURT’S OBSERVATIONS More than 80 % of the payments were interim payments to operational programmes (OPs) of the 2007-2013 programming period, whose eligibility period ended 31 December 2015. The advance payments to the 2014-2020 programming period amounted to around 7,8 billion euro. Brief description of Economic, social and territorial cohesion Policy objectives, instruments and management of spending Spending under the MFF heading 1.b — Economic, social and territorial cohesion aims to reduce development disparities between different regions, restructuring declining industrial areas and to encourage cross-border, transnational and interregional cooperation. The policy instruments that support the reinforcement of economic and social cohesion are the following: (1) the European Regional Development Fund (ERDF), which aims to redress the main regional imbalances through financial support for the creation of infrastructure and productive job-creating investment, mainly for businesses; (2) the Cohesion Fund (CF), which finances environment and transport projects in Member States with a per capita GNP of less than 90 % of the EU average; (3) the European Social Fund (ESF), which aims to improve employment and job opportunities (mainly through training measures), encouraging a high level of employment and the creation of more and better jobs; and (4) other instruments/funds such as the European Neighbourhood Instrument (ENI) that supports cross border cooperation and implementation of political initiatives to bring EU and its neighbors closer and the Fund for European Aid to the most Deprived (FEAD), which provides material assistance to help people out of poverty. Txt. 2. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The ERDF, CF and ESF are governed by common rules, subject to exceptions in the specific regulations of each fund. These funds are implemented through multiannual programmes. The Commission shares management with the Member States. For each programming period, the Commission approves the OPs prepared by the Member States through which projects are financed. Projects are selected by the Member State authorities. During the implementation of the projects, beneficiaries declare the costs incurred to their national authorities, which in turn, certify and declare the expenditure to the Commission. Risks to regularity For ERDF/CF expenditure a main risk is that beneficiaries declare costs which are ineligible according to national eligibility rules and/or the less numerous eligibility provisions in the EU structural funds regulations. ERDF/CF spending is also subject to the generally applicable EU internal market rules (mainly public procurement rules and state aid). In previous years, non-compliance with EU and/or national public procurement rules when awarding contracts has been the main risk to the regularity of ERDF and CF spending. The Commission shares this assessment, as detailed in its Staff Working Documents (Analysis of errors in the Cohesion Policy for the years 2006-2009 (SEC(2011) 1179 of 5 October 2011)). The Commission has continued to take specific actions to mitigate these risks: e.g. exhaustive guidance and additional training provided to managing authorities on the identified risks to increase administrative capacity in these areas; audits targeted to most risky areas; timely implementation of financial corrections, interruption and suspension procedures. Such preventive and corrective actions were brought together with new initiatives under a comprehensive action plan set up in 2013 to improve the implementation of public procurement rules and politically endorsed by the Commission in December 2015. Another action plan to support programme authorities in implementing State aid rules, reviewed and simplified in 2014 through an amended general block exemption regulation and further clarified through the publication of a Notice on the notion of aid by the Commission in May 2016, is also under implementation. For ESF expenditure, the major risk relates to the intangible nature of the investments in human capital, and the involvement of multiple, often small-scale, partners in the implementation of projects. These factors can lead to non-compliance with EU and/or national eligibility rules resulting in ineligible costs being accepted by the systems in place. The Commission has taken specific actions in order to mitigate the risks identified, which include in particular preventive and corrective measures such as guidance, training, simplification, a strict policy on interruptions and suspensions of payments and the timely implementation of financial corrections, when necessary. The Commission is also addressing this risk by actively promoting the use of simplified cost options and by insisting on the importance of first-level checks by the Member States. Furthermore, the Commission updates on an annual basis its audit plan in order to address the most significant risks identified. In implementing OPs, Member States’ authorities face competing priorities. Spending has to be subject to appropriate checks intended to ensure regularity and sound financial management. At the same time, there is an interest in absorbing EU funding, which may in practice militate against the consistent application of effective controls which should prevent ineligible expenditure being reimbursed from the EU budget. This issue becomes more important as the end of the eligibility period approaches, since Member States risk losing the unspent share of the funds initially allocated to them. Appropriate checks have to ensure regularity of all expenditure declared throughout the implementation period and up to closure. Closure documents from Member States are required by 31 March 2017 and will provide additional assurance. The Commission in a timely manner adopted closure guidelines on 20 March 2013 (Decision C(2013) 1573) that were completed and updated on 30 April 2015. The Commission services have organised closure seminars for Member States and have updated their risk assessments and consequently the Audit Strategy 2007-2013 to address any potential risk in view of the last part of the implementation period. The Commission has established a Task Force for Better Implementation of the Cohesion Policy funds in eight Member States. While emphasising full respect of legality and regularity of expenditure, this Task Force assisted the concerned Member States in the closure preparatory process. The same proactive and targeted approach will be applied to support the implementation of Member States’ programmes in 2014-2020, taking account of the lessons learnt. Commission has initiated further simplification measures in 2015 In July 2015, the Commission has set up a High Level Group of Independent Experts on Monitoring Simplification for Beneficiaries of the European Structural and Investment (ESI) funds. This group has been tasked to: (5) assess the uptake of simplification opportunities by Member States including Member States' commitments to reduce the administrative burden for beneficiaries as set out in their partnership agreements covering the 2014-2020 ESI funds; (6) analyse the implementation of simplification opportunities in Member States and regions, taking account of a study on the use of new provisions on simplification during the early phase of the 2014-2020 programme period, and their impact on administrative burden and costs; and (7) identify good practice in matters concerning the reduction of administrative burden on beneficiaries. Txt. 3. THE COURT’S OBSERVATIONS Audit scope and approach Annex 1.1 sets out the key elements of our audit approach and methodology. For the audit of Economic, social and territorial cohesion the following specific points should be noted: (a) We examined a sample of 223 transactions as defined in Annex 1.1, paragraph 7. The sample is designed to be representative of the range of spending under the MFF heading. In 2015, the sample consisted of transactions from 15 Member States. (b) For financial instruments under shared management, we also reviewed their disbursement rates (i.e. the proportion of funds used at the level of final recipients). This was done on the basis of the Commission’s progress reporting for 2014 and our examination of seven ERDF financial instruments in seven Member States. (c) We assessed other elements of internal controls and annual activity reports: (i) the Commission’s supervisory activities of audit authorities in 16 Member States; (ii) the annual activity reports (AARs) included reports of Directorate-General for Regional and Urban Policy and of the Directorate-General for Employment, Social Affairs and Inclusion. Txt. 4. We checked, for a second year, whether and to what extent a performance measurement system was in place to assess whether the completed examined projects have achieved their output and result objectives as set out in the projects’ approval documents. We also verified if these objectives were in line with the OP objectives. This was done for the 149 of the 223 examined projects which were completed at the time of the audit. The results of this exercise are presented in the section on the assessment of projects’ performance. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES REGULARITY OF TRANSACTIONS Annex 6.1 contains a summary of the results of transaction testing. Of the 223 transactions examined, 72 (32 %) were affected by error. On the basis of the 33 errors we have quantified, the estimated level of error is 5,2 %. The Commission notes that the level of error reported by the Court is an annual estimate which takes into account corrections of project expenditure or reimbursements affected by errors detected and recorded before the Court’s audit. The Commission underlines that it is bound by the Financial Regulation, which stipulates, in Article 32(2)(e), that its internal control system should ensure, among other things, adequate management of the risks relating to the legality and regularity of the underlying transactions, taking into account the multiannual character of programmes as well as the nature of the payments concerned. The Commission will continue to exercise its supervisory role, in particular by implementing financial corrections and recoveries at the level that corresponds to the level of irregularities and deficiencies identified. The Commission further notes that given the multiannual character of the management and control systems under Cohesion policy, errors made in 2015 may also be corrected up to and at closure, as illustrated in section 2.4.3 of the respective 2015 AARs of DG Regional and Urban Policy and DG Employment, Social Affairs and Inclusion. The Commission notes that the estimated level of error for 2015 is in line with the error rates presented by the Court for the five last years. This confirms that the level of error for the 2007-2013 programming period remains stable and significantly below the rates reported for the 2000-2006 period. This development derives from an improved regulatory framework including simplification, strengthened management and control systems and the Commission’s strict policy to interrupt/suspend payments as soon as deficiencies are identified, as reported in the 2015 AARs. The Commission will continue to focus its actions on the most risky programmes/Member States and implement corrective measures when needed through a strict policy of interruptions and suspensions of payments up to closure, and to apply strict procedures at closure to exclude any remaining material risk of irregular expenditure. The Commission also notes that the error frequency considerably decreased compared to previous years. For the 2014-2020 period the Commission’s corrective capacity is further strengthened by providing for net financial corrections and therefore reducing the possibility for Member States to reuse funds, for significant deficiencies detected by the Commission. This will be an important incentive for Member States to detect and correct serious irregularities before certifying annual accounts to the Commission. Chapter 1 contains an assessment of the accuracy of the figures for financial corrections presented in the financial statements discussion and analysis of the consolidated EU accounts (see paragraphs 1.37 to 1.43). Chapter 1 of the 2012 annual report explained also how financial corrections are taken into account when calculating the estimated level of error. The Commission notes the Court’s willingness to take into consideration financial corrections applied in its calculation of the estimated level of errors. Such corrections stem from the Members States’ own corrective capacity in some cases — an important pillar of the assurance process — or derive from the Commission’s supervisory actions that may include imposition of flat-rate financial corrections to programmes where it considered that there are remaining serious deficiencies and risks. The Commission refers to its replies under paragraph 1.38. Figure 6.2 presents the extent to which the different types of errors contributed to our estimated level of error for 2015. Economic, social and territorial cohesion — Breakdown of the estimated level of error Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The main sources of error for Economic, social and territorial cohesion are the inclusion of ineligible expenditure in the beneficiaries’ cost declarations and the selection of the wholly ineligible projects followed by infringements of EU and national public procurement rules and state aid rules. Over the last four years, we did not detect any quantifiable error related to the use of simplified cost options (SCOs). For all funds during the 2014-2020 programming period, Member States are obliged to use SCOs for small projects under 50000 euro and can decide to apply SCOs more widely. The Commission considers that the Court’s assessment confirms that simplified cost options are less prone to errors and therefore has actively worked with programme authorities since the introduction of the simplified cost options (SCOs) to progressively extend their use. This has already led to positive results. The Commission continues to actively promote the use of SCOs in the 2014-2020 programming period, where they have been significantly strengthened both in the Common Provisions Regulation and in the ESF specific regulation, in order to reduce the administrative burden on the beneficiaries, promote result-orientation and further reduce the risk of error. The Commission intends to make proposals in 2016 to expand the use of SCOs for all ESI Funds in the Common Provision Regulation. Besides providing extensive guidance on the implementation of SCOs, DG Employment, Social Affairs and Inclusion is currently conducting a second round of simplification seminars in certain priority Member States which have experienced recurring high error rates in the 2007-2013 programming period and have not yet made sufficient use of SCOs. Ineligible expenditure When declaring costs to the Commission, national authorities certify that these costs have been incurred in compliance with a number of specific provisions laid down in EU regulations, national legislation or eligibility rules, specific OP rules, calls for interest, decisions approving projects for co-financing or grant agreements. Ineligible expenditure is the main source of error for both ERDF/CF and ESF. We found that ineligible costs had been declared in 8 % of the transactions examined (see Figure 6.3). These cases account for 55 % of all quantifiable errors and make up approximately 2,4 percentage points of the estimated level of error. The Commission will follow up the cases identified by the Court and propose actions as it deems necessary. The Commission notes that in some cases national or regional rules applied to Cohesion Policy expenditure are more demanding than those foreseen in the EU or the national legislation for similar nationally funded expenditure. These additional requirements can be seen as an instance of unnecessary administrative burden and complexity to Cohesion Policy expenditure imposed by the Member States to themselves and their beneficiaries, as described for example in the reports on gold plating and simplification issued by DG Employment, Social Affairs and Inclusion in 2013. The Commission also notes that three errors occurred because national rules were stricter or more complex than necessary (gold-plating). The Commission estimates that the breaches to these national rules contributed one percentage point to the error rate. Example of ineligible costs declared Ineligible costs declared: An ERDF project in United Kingdom provides financial support to SMEs for consulting services to promote and improve their products or services. However, only a part of the funding granted was given as a grant to SMEs and around 13 % was retained by the beneficiary. The amounts retained do not meet the eligibility conditions for co-financing since this expenditure was not incurred, not paid and not recorded as costs in the accounts of the beneficiary. As a result, a part of the grant is ineligible for co-financing. Ineligible projects We identified four ERDF projects for which the eligibility conditions set out in the regulations and/or the national eligibility rules were not met. These projects account for 12 % of all quantifiable errors making up approximately 1,5 percentage points of the estimated level of error (see Figure 6.4). The Commission will follow up the cases identified by the Court and propose actions as it deems necessary. Example of an ineligible project Ineligible beneficiary: For an ERDF project in the Czech Republic, the call for proposals specified that only SMEs were eligible. The managing authority provided funding to a beneficiary on this basis, which however had not been confirmed at the time the project was selected. Other ineligible projects were also found in the Czech Republic, in Italy and Poland. Breaches of public procurement rules Public procurement rules are a key instrument for spending public money economically and effectively and for establishing an internal market within the EU. Public procurement rules are applicable for all public spending in the Member States and are not specific to Cohesion policy. Non-compliance with EU or national public procurement rules has been over years a major source of errors in this policy area, in particular for regional and urban policy, mainly due to the types of projects co-financed. The Commission has therefore taken various preventive and corrective actions since the last programming periods in order to address weaknesses identified in that area. The Action Plan on public procurement set up in 2013 and politically endorsed by the Commission in December 2015 aims at further improving the implementation of public procurement rules in the Member States through additional preventive measures. The legal framework for ESI Funds 2014-2020 has also introduced ex ante conditionalities for the effective and efficient use of Union funds, which cover inter alia Member States’ public procurement systems. Suspensions of the interim payments may be decided in case of failure to fulfil an applicable ex ante conditionality by the end of 2016 (Article 19 of Regulation (EU) No 1303/2013). The Commission will follow up all errors reported by the Court in accordance with the Commission Decision C(2013)9527 final which sets out guidelines for determining financial corrections to be made by the Commission to expenditure financed by the Union under shared management, for non-compliance with the rules on public procurement. The 2014 Directives on public procurement which introduced simplifications will start producing their effect on the ground once transposed (see also Commission reply to paragraph 6.25). In 2015, we examined 140 public procurement procedures related to contracts for works and services underlying the expenditure for the transactions tested. The estimated contract value for these public procurements amounted to approximately 2,9 billion euro. The vast majority of these contracts are for projects co-financed by ERDF/CF OPs. We identified instances of non-compliance with EU and/or national public procurement rules for 26 of the 140 procedures examined. For four of the cases, a lower number than in previous years, the failures to comply with these rules were so serious that we could determine that the contract should have been awarded to a different bidder or that competition was restricted in such a way that led to an unjustified direct award. We classified these errors as quantifiable (see Figure 6.5). These errors account for 12 % of all quantifiable errors and make up approximately 0,7 percentage points of the estimated level of error. For estimating the level of error we used the provisions of the 2014 public procurement directive which were less stringent than those included in the 2004 public procurement directives (i.e. modifications of contracts). Example of serious failure to comply with public procurement rules Additional works awarded without proper public procurement: For an ERDF road construction project in Germany, additional works in excess of 50 % of the initial contract value were directly awarded to the same contractor. This is a breach of Article 31 of the public procurement directive 2004/18/EC. As a result, these costs are ineligible for EU co-financing. Similar cases were found in other ERDF projects in Italy and the United Kingdom. Other errors relating to tendering and contracting procedures occurred in a further 22 of the 140 public procurements examined. These errors include cases of the incorrect application of the selection criteria, shortcomings in the tender specifications and non-compliance with information and publication requirements. These errors do not contribute to the error rate we estimated. The Commission will follow up the cases identified by the Court and propose actions as it deems necessary. For two transactions in two Member States, the national authorities had identified serious cases of non-compliance with public procurement rules and imposed specific financial corrections at project level. These corrective measures have been taken into account in the calculation of our estimated level of error, since the corrective measures were taken before the notification of our audit. The Commission notes that adequate corrective action is indeed taken by some programme authorities in the quoted Member States and acknowledges that it is reflected by the Court in its quantification of errors. The objective of the Commission’s action plan referred to in the reply to paragraphs 6.19 to 6.21 is to increase the occurrence of such proactive corrective actions at Member State level. On 26 February 2014, the Council of the European Union and the European Parliament adopted three directives aimed at simplifying public procurement procedures and making them more flexible. EU countries had until April 2016 to transpose the new rules into national law (except with regard to e-procurement where the deadline is October 2018). In May 2016, the Commission has addressed letters of formal notice whereby the Commission reminds them of their overdue obligation to enact in national law the three EU directives to 20 out of the 28 Member States. If Member States continue to fail in their obligation to enact in national law, the Commission may open infringement procedures under Article 258 TFEU which could eventually lead to the imposition of fines to the Member States concerned. The Commission is closely monitoring the state of transposition in Member States. As of 30 June 2016, 16 Member States still have to transpose the directive on public procurement (2014/24/EU), while 19 have to transpose the directive on the award of concession contracts (2014/23/EU) and 17 the directive on procurement by entities operating in the water, energy, transport and postal services sectors (2014/25/EU). We have also published a special report on the actions taken by Member States and the Commission to address the problem of public procurement errors in the area of Economic, social and territorial cohesion. The Commission has accepted all recommendations made by the Court in its special report and is taking the recommended actions. The Commission refers to its replies to paragraphs 6.7 to 6.15 and 6.19 above. Infringements of state aid rules State aid is in principle incompatible with the internal market since it may distort trade between Member States. There are, however, some exceptions to this rule. The Commission directly enforces the EU state aid rules. Member States must notify all cases of state aid to the Commission (either through a scheme or case-by-case for a project), unless the project is below a de minimis ceiling or is covered by the General Block Exemption Regulation (GBER). For all cases notified, the Directorate-General for Competition forms a view on whether the aid is compatible with the internal market The ESF is less prone to infringements of state aid rules since very often projects fall under the de minimis rule. In case of a State aid notification, the Commission assesses whether the support constitutes State aid and, if so, whether it is compatible with the internal market. The outcome of the assessment leads to a Commission decision. We identified this year seven ERDF/CF projects in five Member States that infringed the EU state aid rules. Where necessary, we requested and obtained a preliminary assessment by the Directorate-General for Competition. This assessment and the case law of the European Court of Justice were taken into account when classifying errors. The Commission will follow up the cases identified by the Court and propose actions as it deems necessary. For three of these projects, we consider that the project should have obtained less or no public funding from the EU and/or the Member State according to the state aid rules. The main reasons for non-compliance with state aid rules refer to the excess of aid intensity limits These quantified errors make up approximately 0,3 percentage points of the estimated level of error (see Figure 6.6). We applied the 2014 GBER for estimating the level of error in this category. Example of a project infringing state aid rules The aid intensity limit was exceeded: For a CF project in Poland related to the construction of a port infrastructure, funding exceeded the maximum allowable under state aid rules for this type of project. Other cases were found in another ERDF projects in Latvia and the United Kingdom. We have also published a special report on the Commission’s and Member States’ efforts needed to prevent, detect and correct infringements of state aid rules in the area of Cohesion. The Commission refers to its replies to the Court’s observations and recommendations published in the referred special report and will follow up the accepted recommendations. Value-added tax declared by public bodies Value-added tax (VAT) is an EU indirect tax on the domestic consumption of goods and services levied at each stage in the chain of production and borne by the end consumer. According to EU legislation, recoverable VAT is not eligible for co-financing under ERDF/CF and ESF. Non-recoverable VAT can, thus, be declared for co-financing. The regulatory provisions mentioned by the Court for Cohesion policy are the outcome of a policy choice made by the co-legislators. They are overall aligned with the standard principles on eligibility of VAT for grants in other EU policies, in particular under direct management. When public bodies do not declare and therefore do not recover VAT, both the Commission and the Member States have interpreted the regulations as allowing VAT to be an eligible cost. However, the distinction between recoverable and non-recoverable VAT in the regulations is a source of confusion. In principle, the Commission relies on the Member State’s assessment of the beneficiary’s taxable or non-taxable status regarding VAT. However, as confirmed by the Court of Justice in 2012, the notion of recoverable VAT does not result only from the formal examination of the taxable/non-taxable status of the beneficiary. The Commission intends to provide clarification to Member States in that respect for the current 2014-2020 programming period. The Commission will follow up the cases identified by the Court. For example, the Court of Justice ruled in 2012 that VAT must be considered recoverable whenever an operator charges fees with VAT to the ultimate users of the infrastructure. We identified three cases where recoverable VAT was declared by public bodies as eligible cost (see Figure 6.7). Example of ineligible VAT Recoverable VAT is ineligible for EU co-financing: For an ERDF project in Hungary related to the rehabilitation of an unused mine to construct a new open-air event centre the beneficiary, a public municipality, declared VAT as eligible expenditure. However, the operator of the centre will charge fees with VAT to the ultimate users of the infrastructure. As a result, VAT expenditure is ineligible for EU co-financing. Similar cases were found in two ERDF projects in Germany. Even when VAT declared as expenditure by national governments acting as beneficiaries has been considered as eligible, the VAT as such does not represent a net cost to the Member State. A large infrastructure project can be implemented, for example, by a ministry or a body directly under its authority and include VAT in its costs since that ministry does not recover VAT paid. At the same time, the national budget receives the VAT that has been reimbursed by EU funds. The Commission notes that VAT is a general tax collected by national or regional governments which do not redirect it back to individual financed projects. A more stringent position which would declare VAT as ineligible costs for ministries or bodies directly under their authority would actually penalise those which are not in a position to recover VAT or be compensated for it. Such bodies would therefore be obliged to find additional funding to implement their projects or would have to contribute more to their financing than other beneficiaries recovering VAT. This is not the policy choice made by the co-legislators for cohesion policy. In the EU the standard VAT rates applied for most of the projects range from 17 % to 27 %. When VAT declared as non-recoverable by the ministries implementing the projects exceeds the national co-financing rate, the EU reimbursement may even exceed the actual costs incurred for the project, net of VAT. For example, this occurs when the co-financing rate of a project is 85 % or higher and the VAT rate paid by the ministry implementing the project is beyond 20 %. If the VAT rate paid would be 23 %, then the total EU contribution for the project would account for 104,6 % of the actual costs incurred net of VAT. This is particularly relevant for the Cohesion Fund where most of the projects have a co-financing rate of 85 % or higher. Insufficient reliability of checks at Member State level In 18 cases of quantifiable errors made by beneficiaries, national authorities had sufficient information to prevent or detect and correct the errors before declaring the expenditure to the Commission. If all this information had been used, the estimated level of error for this chapter would have been 2,4 percentage points lower. In addition, we found that for two cases, the error that we detected was made by managing authorities and intermediate bodies in Member States. These errors contributed 0,6 percentage points to the estimated level of error. The Commission is strictly following up these cases and agrees that sound and timely management verifications must be in place in order to prevent irregularities occurring in the first place or being included in payment claims. Since 2010, the Commission has been carrying out targeted audits on management verifications of high-risk programmes where it has identified that deficiencies could remain undetected or not detected in a timely manner by the programme audit authority. For DG Regional and Urban Policy and DG Employment, Social Affairs and Inclusion, results of these audits conducted up to 2015 are presented in their respective 2015 AARs (see pages 65 and 56/57, respectively). The Commission refers to the reinforced procedures in the regulatory framework for the 2014-2020 programming period, where management verifications and controls (including on-the-spot checks) have to be carried out on time for the certification to the Commission of programme accounts and submission of management declarations by the managing authorities on an annual basis. Audit authorities have to estimate reliable residual levels of error in the accounts as a result of all verifications, controls and corrections made since the end of the accounting year. The Commission considers that these reinforced control procedures should result in lasting reductions of the error rate. Furthermore, the Commission has developed new guidance in order to further strengthen the reliability of management verifications and to clarify the increased accountability of managing authorities and their intermediate bodies through management declarations and annual summaries in the 2014-2020 programming period. This guidance, which draws on the lessons learned from the previous programming period, was published in September 2015. THE COURT’S OBSERVATIONS EXAMINATION OF FINANCIAL INSTRUMENTS UNDER SHARED MANAGEMENT Financial instruments provide assistance to enterprises or urban projects by way of equity investments, loans or guarantees. They can be used in mainly three areas: for the support of small-medium enterprises (SMEs), for urban development and for the promotion of energy efficiency. By the end of 2014, 1025 financial instruments had been set up under 183 ERDF and ESF OPs in all but three Member States (Ireland, Croatia and Luxembourg). Taken together, they have an endowment of around 16018 million euro. In general, the funds implementing these financial instruments receive a contribution from the OP when their own legal structure is set up, and subsequently use this money to support projects. Such financial support can be provided only to projects which fall within the scope of the OP. Financial instruments are designed to have a revolving character or, for certain types of guarantee funds, to achieve a high leverage effect. Any resources returned from investments or loans made, including profits, are to be used again for the purpose of the activities implemented by the financial instrument. Only disbursements made to final recipients and fund managers’ costs and fees are considered eligible. Unused endowments of financial instruments are to be returned to the EU budget at closure. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Slow implementation of financial instruments The average disbursement rate for 1025 ERDF and ESF financial instruments was 57 % at the end of 2014. This represents a 10 percentage points increase in comparison to 2013 and a 20 percentage points increase in comparison to 2012. The Commission notes that the average disbursement rate of 57 % (at end of 2014) — an increased rate compared to 47 % at the end of 2013 — is an average at EU level that does not reflect the heterogeneous situation at programme level (see the Commission report published on 1 October 2015). The low disbursement rate at the end of 2014 is a result of a number of factors including the financial crisis, the limited experience in some Member States and the late start of some instruments (even in 2014). The Commission would further like to stress that taking into account the life cycle of the financial instruments, an increase is expected in the last year of implementation. Any conclusion on the effective disbursement rates can be fully drawn only at closure. In any event, at closure, initial amounts in financial instruments not disbursed to final recipients will be considered ineligible, recovered by the EU budget and lost for the programmes and Member States. The Commission believes that its report to be published on 1 October 2016 on the situation at end-2015 will show further progress in implementation. Figure 6.8 presents the disbursement rates for financial instruments financed by Economic, social and territorial cohesion as at 31 December 2014. Already in our last annual report we have pointed to the difficulties of several Member States to use the endowments of the financial instruments in full. We note particular problems with low disbursement rates in six Member States (Greece, Spain, Italy, Netherlands, Austria and Slovakia) compared to the EU average. As already mentioned in its reply to the Court’s Annual Report for year 2014 (paragraph 6.49), the Commission has undertaken together with the Member States a number of measures to ensure that the remaining investments made by financial instruments will reach final recipients in a timely manner. The Commission continued its efforts in this respect also in 2015. Disbursement rates for financial instruments financed by Economic, social and territorial cohesion as at 31 December 2014 Source: European Court of Auditors based on EGESIF_15-0027-00, 23 September 2015. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Information on implementation of ERDF and ESF financial instruments is based on data received directly from the Member States. Since 2011, the Commission has made substantial efforts to improve the quality of these data. Nevertheless, the data presented in the Commission’s report are in some cases incomplete or inaccurate. This was also the case for four of the seven financial instruments that we examined in 2015. In May 2015, the Commission provided guidance on reporting instructions to Member States, clarifying the reporting requirements. Furthermore, the Commission has improved its reporting tools for 2014-2020: Member States will have to report the data required under Article 46 of the Common Provisions Regulation which is more comprehensive than the data required for 2007-2013. In April 2016, the Commission services presented a further guidance to Member States on this reporting. Ineligible final recipients In two of the seven financial instruments examined guarantees or loans were approved for final recipients who were ineligible according to the regulation and the national law. These cases are classified as quantifiable errors either under the ineligible expenditure or under the ineligible projects category (see Figure 6.9). The Commission notes that for two errors linked to ineligible recipients in financial instruments, it cannot at that stage assess the impact of these individual errors on the EU budget. As indicated at point 3 of the Guidelines for determining financial corrections to be made to expenditure co-financed by the EU under the structural funds and the European fisheries fund for non-compliance with the rules applicable to financial engineering instruments for the 2007-2013 programming period, individual irregularities can be replaced with legal and regular investments carried out by the financial instruments in final recipients by partial or final closure. The Commission therefore considers that, without contesting the underlying errors, it cannot assess the impact of such errors on the EU budget at this stage. Example of an ineligible loan (ineligible recipient) Ineligible final recipients: In the case of one of the ERDF financial instruments in Hungary, the financial intermediary issued a loan for a company belonging to its own group. However, the regulation and the applicable Commission guidance explicitly forbid that a company belonging to the same group can be considered as an eligible final recipient. As a result the loan is ineligible for co-financing. Other similar cases of ineligible recipients were found in Italy. Extension of the 2007-2013 eligibility period through a Commission decision only We recall a specific issue that was disclosed in the 2014 annual report. Article 56(1) of Regulation (EC) No 1083/2006 established the 31 December 2015 as the end of the eligibility period of payments. During 2015, the Commission decided to extend the eligibility period until 31 March 2017 through a Commission Decision instead of asking the Council and Parliament to amend the Regulation. Since there is still no legislation of sufficient hierarchy we will consider disbursements made after 31 December 2015 to be irregular. The Commission expressed its position on the same observation already in its reply to the Court’s Annual Report for the year 2014. The Commission has acted in line with the European Council’s recommendations of December 2014 and within the margin offered by the existing regulatory framework. Therefore, the Commission considers the disbursements made to final recipients made until end March 2017 as eligible expenditure. The modification decision on the closure guidelines of the Commission does not affect Article 56(1), according to which the contribution from the funds into financial instruments must have been paid at latest at 31 December 2015. This remains the rule. The modification clarifies the Article 78(6) of Regulation (EC) No 1083/2006. Article 78(6) of Regulation (EC) No 1083/2006 is a lex specialis to Article 56(1) of that Regulation. On this basis the Commission concludes that: (13) Article 78(6) implies that the contribution from the operational programme to the financial engineering instrument has to be paid until 31 December 2015, (14) as regards eligibility of expenditure paid by a financial engineering instrument to the final recipient, Article 78(6) allows such expenditure to be eligible until the date of submission of closure documents, (15) as 31 March 2017 is the final date for submission to the Commission of the closure documents, the modification of the closure guidelines clarifies that closure as referred to in Article 78(6) of the General Regulation is the date of submission of the closure documents, i.e. 31 March 2017 (see also Commission reply to paragraph 6.52 and to Recommendation 5 of the Court’s 2014 annual report). Txt. 5. The Commission therefore disagrees that disbursements made after 31 December 2015 are to be considered irregular. We have also published a special report on the lessons to be learnt from the implementation of the EU budget through financial Instruments for the 2007-2013 programming period. The Commission refers to its replies to the Court’s recommendations published in the referred special report and is already taking actions to follow up the accepted recommendations. EXAMINATION OF ELEMENTS OF INTERNAL CONTROL SYSTEMS, INCLUDING ANNUAL ACTIVITY REPORTS Assessment of the Commission’s supervision of audit authorities Commission uses the work of audit authorities in Member States to estimate the level of error Audit authorities provide assurance to the Commission as to the effective functioning of the management systems and internal controls for an OP and the legality and regularity of the expenditure certified. This information is provided by audit authorities in their annual control reports, audit opinions and system audit reports. Since 2009, the Commission has carried out enquiries of the main audit authorities, including on-the-spot reviews of their work, to assess the level of reliance they can place on them. The Commission is closely cooperating and coordinating with audit authorities, and started reviewing their methodologies and audit results as early as 2009. This contributed to capacity building by providing advice, guidance and recommendations to audit authorities through the Commission’s re-performance of audit work carried out by audit authorities. The Commission underlines that its assessment of the reliability of the work of audit authorities is to be seen in that context. As a result of its audit programme that entailed 316 audit missions carried out on the spot cumulatively since 2009 covering 51 audit authorities in charge of auditing around 98,5 % of ERDF/CF allocations for the 2007-2013 period, DG Regional and Urban policy concluded by end 2015 that the work of the reviewed audit authorities can in general be relied upon. Concerning DG Employment, Social Affairs and Inclusion, at the end of 2015, 236 audit missions have been carried out since 2009 covering 89 audit authorities out of 92 (96,7 %) in charge of 115 out of 118 OPs representing 99,6 % of the expenditure. The Commission concluded that it can rely on 82 audit authorities covering 98,2 % of ESF allocations. In their 2015 AARs, DG Regional and Urban Policy and DG Employment, Social Affairs and Inclusion provided a detailed assessment of the accuracy and reliability of the audit information and results reported by audit authorities in their 2015 annual control reports in relation to 2014 expenditure (see pages 55 to 60 of DG Regional and Urban Policy’s 2015 AAR and pages 52 to 54 of DG Employment, Social Affairs and Inclusion’s 2015 AAR). The Directorates-General for Regional and Urban Policy and for Employment, Social Affairs and Inclusion make use of this information when preparing their annual activity reports (AARs) and throughout the year to decide on possible interruptions and/or suspensions of payments to OPs. In 2015, DG Regional and Urban Policy indicated in its AAR that approximately two thirds of the interruptions and pre-suspensions of payments were based on audit results reported to the Commission by audit authorities during the year or at year end (see page 68 of DG Regional and Urban Policy’s 2015 AAR). In the course of 2015, interruptions and pre-suspension procedures concerned 117 ERDF/CF programmes and almost EUR 9,1 billion of payment claims submitted by the Member States but not paid unless the Commission has obtained additional evidence that all necessary corrections had been made by the Member State concerned. For ESF, as reported on page 59 of DG Employment, Social Affairs and Inclusion’s AAR in 2015, the Commission sent 2 warning letters and 17 pre-suspension letters; it decided 27 interruptions of payments and suspended 10 operational programmes. In total, EUR 1,4 billion of payment claims were interrupted. In order to assess whether they can rely on this information, the directorates-general check the error rates reported by the audit authorities for each OP (or group of OPs): (16) if the Commission considers the error rate to be reliable (and representative for the expenditure certified), it accepts the rate reported by the audit authority. Also the Commission may recalculate the rate for its own assessment based on additional information obtained from the audit authority; (17) in the case of unreliable error rates, the Commission applies a flat error rate (between 2 % and 25 %) in line with the results of its assessment of the functioning of management and internal control systems. Txt. 6. The Commission underlines that this assessment of the reliability of error rates each year is based on a thorough desk analysis of all available information completed by risk-based on-the-spot fact-finding missions, also taking account of the overall assessment of the reliability of the work of audit authorities as a result of the comprehensive audit work referred to in the reply to paragraph 6.47. When necessary, the Commission requests and obtains from the audit authorities any additional information required. The Commission also calculates a cumulative residual risk for each OP (or group of OPs), which takes into account all financial corrections since the start of the programming period. This includes corrections already implemented at EU and/or national level. Based on these two indicators and additional information at its disposal, the Commission forms its assessment of the management and control system for the OP. This assessment is then reported in the annual activity report of the directorate-general. The Commission assesses the effective functioning of the management and control system for each programme and for each authority (managing, certifying and audit authorities), based on all available EU and national system audit results on 15 key regulatory requirements according to a methodology shared with audit authorities. All programmes are therefore assessed against audit opinions at national and Commission level based on audits carried out on systems and representative samples of operations. Error rates reported by Member States and residual error rates calculated by the Commission are important elements of this assessment, but not the only ones. Operational line managers and authorising officers by sub-delegation also assess the level of management assurance for each programme. For the purpose of the assurance disclosed in the AARs, the opinion for each operational programme is, therefore, the result of an overall assessment based on all information available to the Commission. This provides a comprehensive and detailed assessment of the effectiveness of management and control systems per Member State and, for operational programmes included in the 2015 reservations, information on the main issues identified and the actions taken. Commission’s assessment of annual control reports can only partially address the risk of unreliable information being presented by national authorities The robustness of the Commission’s assessment of ERDF/CF and ESF OPs depends on the accuracy and reliability of information reported by Member State authorities. Our audits in previous years identified two main risks in this regard: (18) audit authorities may understate the errors and/or may not extrapolate them properly. As a consequence, the reported error rates may not always be reliable; (19) information on financial corrections reported by Member States may not always be reliable or accurate and, as a result, the Commission’s calculation method may result in an understated cumulative residual risk. Txt. 7. As explained in their respective AARs, both directorates-general have taken measures to improve the reliability of audit information (error rates and audit opinions) and financial corrections reported by the certifying authorities. To mitigate the risks mentioned by the Court, the Commission is implementing two audit enquiries: (20) the Commission reliance on the audit information, reported by audit authorities, including the annual control reports, the audit opinions and error rates, is based on wide desk and on-the-spot audit work. This aims at verifying whether audit findings are confirmed, including through on-the-spot re-performance by Commission auditors. The desk review of annual control reports covers a wide range of verifications linked e.g. to the accuracy of the calculations, parameters used, treatment of errors detected or reconciliation of the audited population and is complemented by fact-finding missions (see also the Commission’s reply in paragraph 1.36 and 6.47). Additional specific information is requested and obtained from audit authorities when necessary in case of doubts but is not systematically requested in all cases, in line with the single audit principle and the regulatory requirements. As a result of this in-depth audit work, Directorates-General for Regional and Urban Policy and Employment, Social Affairs and Inclusion have assessed 95 % of the Member States error rates for their policy areas as a reliable source of information for the purpose of calculating the risk to 2015 payments (pages 58 and 53 of respective 2015 AARs). For the remaining 5 % where the error rates reported by the Member States were considered unreliable the Commission replaced them by a best estimate on the basis of flat rates, (21) specific risk-based on-the-spot audits (so far covering 75 OPs over the last three years) in order to ensure that the corrections reported are effectively implemented and, in case of doubts or insufficient evidence, deduction of the amounts concerned from the cumulative financial corrections taken into account for the purposes of the calculation of the residual error rate. As a result of their audits and desk consistency checks on the reliability of the article 20 statements on financial corrections submitted for all programmes, the Commission services decided to take a prudent approach and exclude some of the reported financial corrections for the calculation of the cumulative residual risk (see the annual activity report of Directorate-General Regional and Urban Policy, pages 72 and 73). Txt. 8. Contrary to 2007-2013, in the 2014-2020 programming period audit authorities will issue their audit opinion based on residual error rates after corrections have been implemented. As a result, the accuracy of financial corrections will be systematically audited as part of the audit on accounts. The Commission’s verifications of the annual control reports consist mainly of desk reviews and can therefore only partly address these risks. In particular, we consider that the Commission’s scope for validating (and, where necessary, adjusting) the reported error rates is limited when audit authorities are not requested to provide the Commission with more specific information on their audits of operations (for example details on the audit scope, coverage/sub-sampling and error classification) for its verification of the annual control reports. Our analysis showed that in 2015 such information has been requested more often than in previous years. During 2015, as in previous years, both directorates-general also performed on-the-spot checks on financial corrections implemented by ten Member States. For seven of these, the Commission identified deficiencies requiring adjustments to the financial corrections reported. In 2014, the Commission detected deficiencies for seven of twelve Member States visited. These findings confirm our own assessment and highlight, in the context of the new 2014-2020 programming period, the importance of ensuring that Member States put in place reliable reporting systems. Commission considers that expenditure declared for 58 % of all OPs was free of a material level of error For 2015, the Commission considers that it had assurance that 254 of the 440 ERDF/CF and ESF OPs (58 %) were free from a material level of error: audit authorities had reported error rates below the Commission’s materiality threshold of 2 %, and these rates had been validated by the Commission. These OPs represent around 38 % of the 2015 payments. In 2015, as in previous years, we examined whether the Commission was effective in verifying the accuracy and reliability of the error rates reported by audit authorities. This was carried out for a sample of 102 OPs in 16 Member States (80 of the 322 ERDF/CF OPs and 22 of the 118 ESF OPs) on the basis of the Commission’s working files. Our sample included OPs for which the Commission had accepted the rates reported by audit authorities, but also OPs for which rates had been recalculated or where flat rates had been imposed. Commission’s validation (or recalculation) of the error rates is consistent with the evidence provided by audit authorities In relation to five OPs covered by five ACRs, we identified weaknesses that had not been detected or properly addressed by the Commission. However, we consider that none of these cases should have led the Commission to make additional reservations (or reservations with a higher financial impact) in the 2015 annual activity reports. This is so since almost all of these OPs (or group of OPs) are already under reservation for different reasons. The Commission notes the improvement in the assessment of its verification of errors rates reported by the audit authorities. Improvements in audit authorities’ checks on state aid and public procurement As in previous years, we also examined whether national audit authorities carried out appropriate checks of public procurement and state aid rules. We noticed a significant improvement in that area compared to the previous year’s assessment since most of the check lists used by audit authorities in their audit of operations adequately covered compliance with state aid and public procurement rules. As a result, for two of the 102 OPs sampled the state aid checks carried out by the audit authorities were inadequate (for example providers of services of general economic interest were not covered or for notified aid schemes and research and development measures the aid intensity was not verified). For one of these two OPs the public procurement checklist is considered insufficient. Through its own verifications, the Commission also noted the significant improvement reported by the Court. During the ACR review, the Commission examines whether the audit authority’s opinion is solidly grounded on the basis of the results of system and operations audits. The examination of the work of the audit authorities with regards to state aid and public procurement is included, amongst other issues, in the Commission’s review on audit authorities which include an examination of the methodology for system audits and audits of operations (including checklists) as well as re-performing audits already done by audit authorities (see reply to paragraph 6.47). In some cases the Commission had already recommended to the audit authorities concerned, through its own audit reporting, to reinforce their checklists to better cover the verification of state aid and public procurement issues. The Commission will continue to ensure that audit authorities use appropriate checklists. Commission guidance requires audit authorities to include contributions to financial instruments to the audited population rather than checking the actual disbursements to final recipients Financial contributions from the OPs to funds implementing financial instruments for the whole period are generally made through a single payment as soon as the legal structure of the funds has been set up. Alternative financing arrangements can be specified in the funding agreements. In accordance with regulation, these payments are certified to the Commission as expenditure incurred. In line with the Commission’s guidance to audit authorities on sampling, ex post verification of such payment by audit authorities generally relate to the financial year in which the payment has been made. These checks generally have a limited scope because within the first year few operations are selected for support and limited amounts are disbursed to final recipients and projects. In line with Article 78(6) of Regulation (EC) No 1083/2006, the funds’ contribution into financial instruments is included in payment claims to the Commission. These payments constitute part of the population to be audited by audit authorities through sample checks in line with the updated Commission guidance. The audit framework developed in 2011 for ERDF/CF and ESF clearly foresees that audits should cover the set-up and the implementation of the financial instrument and that specific verifications by closure are envisaged. Audit authorities and Commission may carry out thematic audits on financial instruments which cover both the set-up of the fund and the actual implementation of projects through the audit of a sample of projects carried out by final recipients. By letter of 18 March 2015, the Commission has also clarified to the audit authorities that financial instruments should always be included in the population subject to sampling and has informed them that it considered good practice that these operations are audited as a separate stratum for the purposes of sampling. In addition, during the Homologues Group Meeting of September 2015, the Commission discussed with audit authorities the audit approach and audit strategy towards closure. The necessary assurance on financial instruments at closure will be obtained by combining results from system audits carried out during the implementation, audits of a sample of transactions which covered the implementation when this was possible and additional testing by closure to ensure that actual disbursements to final recipients are legal and regular. This discussion was systematically followed up in early 2016 in bilateral audit coordination meetings with the concerned Member State’s audit authorities. Finally, for the purpose of the closure declaration, the audit authorities will have to obtain and provide assurance on the legality and regularity not only of the OP contribution paid into financial instruments but also of the use of this contribution through underlying transactions implemented by the final recipients as listed in Article 78.6 of Regulation (EC) No 1083/2006. For three OPs the contribution to financial instruments accounts for a considerable part of the the population and of the audited sample. The audit authorities found these contributions to be free of error. Unless specific verifications are envisaged at closure, the actual implementation of financial instrument operations is unlikely to be checked by audit authorities before closure. In the last two years, we detected the selection of ineligible recipients in some of the instruments audited (see also paragraph 6.44). The Commission has not issued a formal horizontal guidance on the audit of the disbursements made from the financial instrument to final recipients up to closure. However, the audit approach is agreed upon bilaterally between the Commission and the Member State concerned. Where the financial instruments are managed by the European Investment Bank (EIB), audit authorities cannot check the contributions to the fund due to limitations imposed to access rights by EIB. Further discussions between the Commission and the EIB Group should allow managing and audit authorities to have assurance at closure on this specific point. Very low error rates reported by the audit authorities for OPs for which we have found errors For 110 of the 440 ERDF/CF and ESF OPs, the Commission has validated error rates below or equal to 0,5 %. For 16 of these OPs, we performed at least one examination during the last two financial exercises. For nine OPs we reported quantifiable errors in the projects examined (see paragraphs 6.55 and 6.56). The Commission notes that the fact that individual errors are found in a programme does not prejudge audit results based on statistical samples adequately and randomly selected or samples covering a significant part of the expenditure declared. The Commission considers that individual errors reported by the Court do not put in doubt the reliability of the reported and validated error rates in 2015. Examination of the Commission’s annual activity reports We assessed the 2015 annual activity reports (AARs) and accompanying declarations of the Directorates-General for Regional and Urban Policy, and Employment Social Affairs and Inclusion. In particular: (22) we checked the consistency and accuracy of the Commission’s calculation of the amounts at risk; (23) we assessed the reservations made for 2015. Txt. 9. Commission’s calculation of the amounts at risk for ERDF/CF and ESF OPs Overall, the Commission’s calculation of the amounts at risk in the 2015 annual activity reports is accurate and consistent with the available information reported and/or provided by the audit authorities. We recall however that the annual error rates reported by the Commission in the annual activity reports are not directly comparable to those estimated by us. The Commission agrees that the Court’s error rate and that of the Commission are not directly comparable. However, the objective of this process is essentially the same, i.e. assessment of the risk to the EU budget in a particular year. The Commission takes into account all the differences quoted by the Court in its special report 16/2013 (see Commission reply to paragraph 11 of this report). Except for these differences, the Commission considers that for the 2015 annual report, as was the case for the last five years in a row for Directorate-General for Employment, Social Affairs and Inclusion and four years in a row for Directorate-General for Regional and Urban policy before this annual report, the result of the Commission’s assessment is in line with the error rates calculated by the Court (see pages 65 and 100 of the respective AARs). (24) the Directorate-General for Regional and Urban Policy estimated that between 3,0 % and 5,6 % of the interim and final payments for ERDF/CF OPs of the 2007-2013 programming period authorised during the year were at risk of error. 63 % of payments made by the Commission were to OPs with a validated error rate above 2 %; Txt. 10. Directorate-General for Regional and Urban policy made in 2015 63 % of payments to programmes it assessed at year-end as affected by a material level of error above 2 %, but for the vast majority of these programmes sufficient corrections were made in order to bring the cumulative residual risk below materiality. See also Commission reply to paragraph 6.67. (25) the Directorate-General for Employment, Social Affairs and Inclusion estimated that between 3,0 % and 3,6 % of the interim and final payments for ESF OPs of the 2007-2013 programming period authorised during the year were at risk of error. 56 % of payments made by the Commission were to OPs with a validated error rate above 2 %. Txt. 11. Directorate-General for Employment, Social Affairs and Inclusion made in 2015 56 % of payments to programmes it assessed at year-end as affected by a material level of error above 2 %, but for the vast majority of these programmes sufficient corrections were made in order to bring the cumulative residual risk below materiality. See also Commission reply to paragraph 6.67. Commission’s reservations for ERDF/CF and ESF OPs In cohesion policy, the Commission issues reservations on OPs (or groups of OPs) when: (26) significant weaknesses in the OPs management and control systems have been identified, or (27) the OPs validated error rate exceeds 5 %. Txt. 12. For OPs with a validated error rate between 2 % and 5 %, the Commission also considers whether the OPs cumulative residual risk exceeds 2 %. The Commission notes that as reported in the respective 2015 AARs: (28) 93 % of ERDF/CF programmes presented a cumulative residual risk below 2 % at the end of 2015 (page 58). In relation to the 21 programmes with a cumulative residual risk above 2 %, reservations were made for 20 programmes and for the remaining programme the exception was disclosed in the AAR (page 180), (29) 90,7 % of the ESF programmes, representing 93,3 % of the 2015 payments, presented a CRR cumulative residual risk below 2 % (page 61). Reservations were made for the remaining programmes, unless the necessary corrective measures had already been implemented. Txt. 13. The number of OPs subject to reservation decreased in 2015 from 77 to 67 for ERDF/CF OPs and from 36 to 23 OPs for ESF. Meanwhile, the estimated financial impact of these reservations decreased from 234 million euro in 2014 to 231 million euro in 2015 for ERDF/CF and from 169,4 million euro in 2014 to 50,3 million euro in 2015 for ESF. The Court’s observation demonstrates the maturity of the methodology used for reservations in the respective AARs and the strict implementation of this methodology by DG Regional and Urban Policy and DG Employment, Social Affairs, and Inclusion. Both directorates-general follow a strict approach with regard to the interruption and suspension of payments. They also apply a preventive approach as regards payments, i.e. no payments are made to operational programmes for which an audit mission is planned until the audit mission is completed. Payments are resumed afterwards unless significant deficiencies were identified. We found that the reservations made by the Commission in the 2015 annual activity reports were in line with the Commission’s instructions and the information made available to the two directorates-general. The Commission estimates the cumulative residual risk to be 1 % of the payments for all ERDF/CF OPs and 0,9 % for all ESF OPs. These figures are based on the reported error rates by audit authorities and validated by the Commission and take also into account the financial corrections reported by certifying authorities to the Commission. Looking forward: 2014-2020 programming period In our 2014 annual report we signalled delays in the start of the 2014-2020 programming period. We note that at the end of 2015, fewer than 20 % of the national authorities responsible for ESI funds have been designated. While Member State authorities may for example launch calls and select projects, they cannot submit statements of expenditure to the Commission until their designation. As a consequence, there is a risk that delays in the budgetary execution for the 2014-2020 programming period will be greater than those for the 2007-2013 period. The Commission notes that designation is a Member States’ responsibility. The Commission has encouraged Member States on several occasions (in meetings and written communication) to speed up this process — while ensuring full respect of the designation criteria defined in the regulation — and provided clarifications and assistance where necessary. The Commission further notes that implementation on the ground can start well before the designation process is finalised at the level of the Member States and notified to the Commission. On average already around 10 % of the allocations have been committed to selected operations, with variations between Member States. CONCLUSION AND RECOMMENDATIONS The conclusion for 2015 Overall audit evidence indicates that spending on Economic, social and territorial cohesion is affected by a material level of error. The Commission notes that the estimated level of error presented by the Court remains stable and is also consistent with the error rates reported in the respective AARs, subject to the differences set out in paragraph 6.66. The Commission will continue to focus its audits and actions on the most risky programmes/Member States and implement corrective measures when needed through a strict policy of interruptions and suspensions of payments up to closure, and to apply strict procedures at closure to exclude any remaining material risk of irregular expenditure. The Commission further notes that given the multiannual character of the management and control systems under Cohesion policy, errors made in 2015 may also be corrected in subsequent years and before closure as illustrated in section 2.4.3. of the respective 2015 AARs of DG Regional and Urban policy and DG Employment, Social Affairs and Inclusion. For the 2014-2020 period, reinforced requirements on legality and regularity include the introduction of annual accounts, reinforced management verifications and accountability as well as audit opinions based on residual levels of errors each year, as well as a retention of 10 % on all Commission interim payment until all controls are performed at Member State level and annual accounts submitted. Residual level of errors will be calculated for each accounting year covering the accounts. The Commission’s corrective capacity is further strengthened by the possibility to apply net financial corrections for serious deficiencies under certain conditions, thereby reducing the possibility for Member States to reuse funds. This will be an important incentive for Member States to detect, to report and to correct serious deficiencies before certifying annual accounts to the Commission. All these elements should contribute to a lasting reduction in the error rate. See also Commission reply to paragraph 6.11. For this MFF heading area, testing of transactions indicates that the estimated level of error present in the population is 5,2 % (see Annex 6.1). Our audits since 2009 have shown that the level of error for the 2007-2013 programming period is significantly lower than for the 2000-2006 programming period. It remains, however, persistently above the materiality threshold. We also note that expenditure in this area is reimbursed based on declarations of actual costs incurred by beneficiaries. Several attempts at simplification have been made in previous years, but had only a limited impact on the regularity of transactions. At the same time, increased levels of control have caused a significant administrative burden on national administrations and beneficiaries. In 2015, the Commission set up a high level group of simplification in view of preparing for the post-2020 programming period (see paragraph 6.8). The level of error for the 2007-2013 programming period remains stable and significantly below the rates for the 2000-2006 period. This development derives from an improved regulatory framework including simplification, strengthened management and control systems and the Commission’s strict policy to interrupt/suspend payments as soon as deficiencies are identified, as reported in the 2015 AARs. Moreover, a study commissioned by the Commission in 2010 on the administrative costs of the implementation and management of the funds shows that their share of the total eligible ERDF/CF expenditure is not excessive as it was estimated to 3,2 % for the entire period (approximately EUR 12,5 billion). A further study in 2012 which was used as the basis for the impact assessment for the 2014-2020 period estimated that Commission proposals would lead to a reduction of almost 20 % in administrative burden for beneficiaries. The Commission has also been actively working with the Member States on the promotion of the simplified cost options, which are focused on results and based on calculated costs as opposed to reimbursement of actual expenditure incurred in order to progressively extend their use and considers that these efforts have already led to positive results (see also Commission replies to paragraph 6.15). In order to better understand the barriers to use of simplification measures, the Commission is undertaking a study which should enable it to assess the uptake of simplified cost options on Member State level and the potential for their increased use. The Commission has set up the High-level Group on simplification in order to make recommendations to maximise the possibilities for reducing administrative burden under the 2014-2020 regulatory framework. Recommendations Annex 6.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, we presented 16 recommendations. Of these recommendations, the Commission fully implemented nine recommendations, four were implemented in most respects, two were implemented in some respects, and one was not implemented. The Commission notes that it had not accepted the recommendation reported under this paragraph as not implemented. For further details, see Commission reply to Annex 6.2. Following this review and the findings and conclusions for 2015, we recommend that the Commission: (30) Recommendation 1: should fundamentally reconsider the design and delivery mechanism for the ESI funds when making its legislative proposal for the next programming period taking also into account the suggestions of the high level simplification group (target implementation date: final results before the Commission presents its proposal for the post-2020 period); Txt. 14. The Commission partially accepts the recommendation. While the Commission is not in a position at this stage to make any specific commitment in relation to legislative proposals for the post-2020 period, it will consider all viable options taking into account the suggestions of the High-Level Group as well as results of studies and evidence from 2014-2020 programme implementation and other relevant information. (31) Recommendation 2: should make use of the experience acquired in the 2007-2013 programming period and report on a focused analysis of the national eligibility rules for the 2014-2020 programming period. Based on the above, it should provide guidance to Member States on how to simplify and avoid unnecessarily complex and/or burdensome rules that do not add value with respect to the results to be achieved by the policy (gold-plating), in particular when they result in significant and/or recurring errors (target implementation date: end of 2018); Txt. 15. The Commission accepts this recommendation. It will pay specific attention to national eligibility rules in its monitoring and audit of national management and control systems, in particular in Member States with persistently high error rates On this basis, it will help the Member States concerned to simplify and avoid unnecessarily complex and burdensome rules, facilitate the exchange of good practices and will report on the experience acquired in 2018. The Commission continues to heavily invest on the use of simplified cost options during the 2014-2020 programming period. The Commission recalls that, under shared management, it is primarily the responsibility of the Member States to review and simplify their national eligibility rules. (32) Recommendation 3: should submit a legislative proposal to amend, through a legislative act of equal legal value, Regulation (EC) No 1083/2006 with respect to the extension of the eligibility period for financial instruments under shared management to the Council and the Parliament. A change of the regulation is required to provide legal certainty to the Member States (target implementation date: immediately); Txt. 16. The Commission does not accept the recommendation. As it has stated in its reply to the same recommendation in the 2014 Annual report of the Court of Auditors, it considers that, within the mandate given by the European Council in December 2014, the modifications introduced in its closure guidelines were within the scope of Article 78(6) of Regulation (EC) No 1083/2006, as amended, and therefore did not require an amendment of the legislative act (see also Commission common reply to paragraph 6.59 to 6.62 above and to Recommendation 5 of the Court’s 2014 annual report). The Commission notes that the Council and the European Parliament took note of, but did not support, the recommendation made by the Court in the context of the 2014 discharge. (33) Recommendation 4: for the 2014-2020 programming period, the Commission should clarify to Member States the notion of recoverable VAT, in particular for public beneficiaries, to avoid different interpretation of the term non-recoverable VAT and avoid a sub-optimal use of EU funds (target implementation date: mid 2017); Txt. 17. The Commission accepts this recommendation. It will provide guidance to Member States to clarify the notion of recoverable VAT in the context of the 2014-2020 programming period, within the policy choice of the co-legislators for value added tax (VAT) eligibility, as provided for in Article 69(3)(c) CPR. (34) Recommendation 5: should ensure that all the expenditure related to ERDF and ESF financial instruments for the 2007-2013 programming period are included sufficiently early in the closure declarations to enable audit authorities to carry out their checks. In addition, the Commission should encourage all Member States that implemented financial instruments to carry out specific audits on the implementation of these instruments in view of the closure (target implementation date: immediately). Txt. 18. The Commission accepts the recommendation. It already clarified in its closure guidelines for 2007-2013 that in order for the audit authorities to have sufficient time to carry out their work for the closure declaration, the application for payment of the final balance and the final statement of expenditure should be submitted to the audit authorities well in advance, at a date to be agreed between authorities for each programme. For the purpose of the closure declaration, the Commission has already encouraged audit authorities to obtain and provide assurance on the legality and regularity not only of the programme contribution paid into financial instruments but also of the use of this contribution through underlying transactions implemented by the final recipients. The Commission will on the basis of the closure packages submitted by 31 March 2017 assess if sufficient audit work was carried out on financial instruments as the eligibility of underlying expenditure is determined at closure for such instruments. THE COURT’S OBSERVATIONS ASSESSMENT OF PROJECTS’ PERFORMANCE Implementing the EU budget under the principle of sound financial management requires that spending should not only focus on complying with the rules but also on the achievement of intended objectives. 149 of the 216 projects we examined were completed at the time of the audit. All 15 Member States sampled had at least three completed projects that could be assessed (see paragraph 6.9). In addition to checking the regularity of these 149 transactions, we assessed for a second year, based on information provided by the beneficiaries, whether and to what extent: (35) a performance measurement system was in place to assess whether the objectives specified for projects in the project’s approval documents (project application, grant agreement, contract and/or decision for co-financing) were in line with those set out in the OPs and whether appropriate result indicators were set in order to measure the projects’ performance; (36) projects had achieved those objectives and met the targets set for each indicator. Our analysis was broader for this year compared to the previous year as we did not focus only on project outputs, but also on results in order to measure the projects’ performance in contributing to the OP result objectives. Txt. 19. Figure 6.10 summarises the results on the assessment of the completed projects’ performance. The results on the assessment of the completed projects’ performance Performance assessment No of projects for which: Both output and result indicators were defined and assessed Outputs were defined and assessed but result indicators were not defined Neither outputs nor result indicators were defined and could not be assessed TOTALS Number of projects 90 57 2 149 (37) Fully achieved Txt. 20. 52 (58 %) 50 (88 %) — 102 (68 %) (38) Partially achieved Txt. 21. 34 (38 %) 7 (12 %) — 41 (28 %) (39) Not achieved Txt. 22. 4 (4 %) — — 4 (3 %) (40) Cannot be assessed Txt. 23. — — 2 (100 %) 2 (1 %) THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES For all but two of the projects we examined there was a performance measurement system in place linking the project to the OP objectives For 147 of the 149 projects (99 %) we conclude that there was a performance measurement system in place based on which an assessment of whether the output indicators are in line with the operational programme indicators can be made (as specified for each priority axis). For the remaining two cases (1 %), we could not assess the output of the projects either because specific targets in the output objectives were not set at the stage of the approval of the project or the managing authority/intermediate body could not provide appropriate documentation to us to substantiate that such targets had been specified and communicated to the beneficiary. Therefore, we cannot assess whether or to what extent these projects contributed to the achievement of the OP objectives. Almost all of the 90 projects with both output and result objectives defined had fully or partially met them For the 2007-2013 programming period there was a legal requirement obliging beneficiaries to define and report on outputs but not on results. Performance had to be monitored at the level of the priority axes of the OP and not at project level. Although the structural funds legislative framework for 2007-2013 did not oblige national authorities to define result indicators at project level, the Commission encouraged this whenever relevant. Some result indicators could indeed not be meaningfully measured at the level of a single project. However, the Member State authorities still had the possibility to define result indicators. This allows linking the projects’ results to the priority axis and determining the contribution of the projects to the achievement of the objectives of the operational programme. Result indicators were defined for 90 of the 149 (60 %) completed ERDF/CF and ESF projects. 52 of the 90 completed projects (58 %) which could be assessed at both output and result objectives level had fully achieved both the output and the result objectives set by the managing authorities (see Figure 6.11). Another 34 (38 %) of these 90 projects had partially achieved their output and result objectives. The Commission notes that 86 projects out of the 90 examined which have defined output and result indicators have reached (either fully or partially) or exceeded their corresponding targets. The Commission notes that, in the first place, it is the Monitoring Committees that are required to ensure that projects with EU added value are selected and approved. Performance evaluation of programmes will be reinforced for the 2014-2020 period: as foreseen in the 2014-2020 regulatory framework (Article 22(6) and (7) of Regulation (EU) No 1303/2013), the Commission will be able to sanction Member States in case of serious underachievement of a priority axis in a programme, as a result of the performance review (Article 22(6) of Regulation (EU) No 1303/2013) and at closure (Article 22(7)). Example of a project which fully achieved its objectives The project met its output and result objectives: An ESF project in Romania consisted of the organisation and delivery of training courses in the field of construction in order to provide participants with new skills that would improve their employment prospects. The grant agreement of the project defined five output and eight result indicators which were in line with the OP indicators at priority axis level. The project achieved all indicators within the deadline set in the grant agreement. The project even exceeded its targets in terms of participation of long-term unemployed women and youngsters. The remaining four projects (4 %) did not achieve any of their objectives. One of those four projects was withdrawn by the national authorities following our audit (see Figure 6.12). The Commission uses all possible measures at its disposal to ensure sound and efficient financial management in the use of EU funds. Moreover, it is only at the stage of the closure of the programme that the performance of projects financed by Operational Programmes will be evaluated. Example of a project which did not achieve its objectives For an ERDF project in Poland related to the development of a new method for burning brown coal, the beneficiary’s technical capability to undertake the project relied on the involvement of external contractors. Some of these contractors did not finally participate to the project. This contributed to the project not achieving its intended objectives. The intermediate body decided to cancel the project following our audit. Result indicators were not defined for 38 % of the projects For 57 of the 149 (38 %) completed projects the Member State authorities implemented a performance measurement system to allow the monitoring of project outputs but did not define any result indicators or targets in the project’s approval documents. For these projects, we could not measure if and to what extent they contributed to the achievement of the OP result objectives. The Commission and Member States assess to which extent OP result objectives have been achieved primarily through ex post evaluations, building on the monitoring and reporting of a series of output and result indicators. Our assessment in 15 Member States sampled showed that such cases were identified in 13 of them. However, the issue is particularly prevalent in Czech Republic, Spain, France, Italy, Latvia and Austria for ERDF/CF and Czech Republic, Germany, Spain and Poland for ESF. In these countries result indicators were not defined for more than half of the completed projects examined. However, 50 of these 57 completed ERDF/CF and ESF projects (88 %) had fully achieved their output objectives set by the managing authorities. The remaining seven (12 %) projects had partially achieved their output objectives. As regards the 2014-2020 programming period, we note that there is no legal requirement in the common provision regulation for Member States to define result indicators at project level. Nevertheless, we consider it good practice that managing authorities and intermediate bodies should whenever possible define relevant result indicators at project level that measure the contribution of the project to the achievement of the objectives set for the priority axis of the OP. The Commission notes that there is a major improvement for 2014-2020 with a comprehensive system of obligatory indicators for outputs and results at programme level. The achievement of the objectives is measured by a result indicator (with a baseline — the starting point — and target). Programmes will systematically measure and report progress of result indicators. Projects within a programme are selected to deliver outputs and contribute towards the results to be achieved at programme level. Their progress is measured with output indicators. These are also mandatory.

THE COURT’S OBSERVATIONS INTRODUCTION This chapter presents our findings for Economic, social and territorial cohesion (MFF heading 1.b). Key information on the activities covered and their spending in 2015 is provided in Figure 6.1. MFF heading 1.b — Key Information 2015 (billion euro) Source: 2015 consolidated accounts of the European Union. Total payments for the year 51,2 - advances 9,1 + clearings of advances 7,4 + disbursements to final recipients from financial instruments under shared management 2,1 + advances used by final beneficiaries of state aid projects 2,3 Audited population, total 53,9 THE COURT’S OBSERVATIONS More than 80 % of the payments were interim payments to operational programmes (OPs) of the 2007-2013 programming period, whose eligibility period ended 31 December 2015. The advance payments to the 2014-2020 programming period amounted to around 7,8 billion euro. Brief description of Economic, social and territorial cohesion Policy objectives, instruments and management of spending Spending under the MFF heading 1.b — Economic, social and territorial cohesion aims to reduce development disparities between different regions, restructuring declining industrial areas and to encourage cross-border, transnational and interregional cooperation. The policy instruments that support the reinforcement of economic and social cohesion are the following:

the European Regional Development Fund (ERDF), which aims to redress the main regional imbalances through financial support for the creation of infrastructure and productive job-creating investment, mainly for businesses;

the Cohesion Fund (CF), which finances environment and transport projects in Member States with a per capita GNP of less than 90 % of the EU average;

the European Social Fund (ESF), which aims to improve employment and job opportunities (mainly through training measures), encouraging a high level of employment and the creation of more and better jobs; and

other instruments/funds such as the European Neighbourhood Instrument (ENI) that supports cross border cooperation and implementation of political initiatives to bring EU and its neighbors closer and the Fund for European Aid to the most Deprived (FEAD), which provides material assistance to help people out of poverty.

THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The ERDF, CF and ESF are governed by common rules, subject to exceptions in the specific regulations of each fund. These funds are implemented through multiannual programmes. The Commission shares management with the Member States. For each programming period, the Commission approves the OPs prepared by the Member States through which projects are financed. Projects are selected by the Member State authorities. During the implementation of the projects, beneficiaries declare the costs incurred to their national authorities, which in turn, certify and declare the expenditure to the Commission. Risks to regularity For ERDF/CF expenditure a main risk is that beneficiaries declare costs which are ineligible according to national eligibility rules and/or the less numerous eligibility provisions in the EU structural funds regulations. ERDF/CF spending is also subject to the generally applicable EU internal market rules (mainly public procurement rules and state aid). In previous years, non-compliance with EU and/or national public procurement rules when awarding contracts has been the main risk to the regularity of ERDF and CF spending. The Commission shares this assessment, as detailed in its Staff Working Documents (Analysis of errors in the Cohesion Policy for the years 2006-2009 (SEC(2011) 1179 of 5 October 2011)). The Commission has continued to take specific actions to mitigate these risks: e.g. exhaustive guidance and additional training provided to managing authorities on the identified risks to increase administrative capacity in these areas; audits targeted to most risky areas; timely implementation of financial corrections, interruption and suspension procedures. Such preventive and corrective actions were brought together with new initiatives under a comprehensive action plan set up in 2013 to improve the implementation of public procurement rules and politically endorsed by the Commission in December 2015. Another action plan to support programme authorities in implementing State aid rules, reviewed and simplified in 2014 through an amended general block exemption regulation and further clarified through the publication of a Notice on the notion of aid by the Commission in May 2016, is also under implementation. For ESF expenditure, the major risk relates to the intangible nature of the investments in human capital, and the involvement of multiple, often small-scale, partners in the implementation of projects. These factors can lead to non-compliance with EU and/or national eligibility rules resulting in ineligible costs being accepted by the systems in place. The Commission has taken specific actions in order to mitigate the risks identified, which include in particular preventive and corrective measures such as guidance, training, simplification, a strict policy on interruptions and suspensions of payments and the timely implementation of financial corrections, when necessary. The Commission is also addressing this risk by actively promoting the use of simplified cost options and by insisting on the importance of first-level checks by the Member States. Furthermore, the Commission updates on an annual basis its audit plan in order to address the most significant risks identified. In implementing OPs, Member States’ authorities face competing priorities. Spending has to be subject to appropriate checks intended to ensure regularity and sound financial management. At the same time, there is an interest in absorbing EU funding, which may in practice militate against the consistent application of effective controls which should prevent ineligible expenditure being reimbursed from the EU budget. This issue becomes more important as the end of the eligibility period approaches, since Member States risk losing the unspent share of the funds initially allocated to them. Appropriate checks have to ensure regularity of all expenditure declared throughout the implementation period and up to closure. Closure documents from Member States are required by 31 March 2017 and will provide additional assurance. The Commission in a timely manner adopted closure guidelines on 20 March 2013 (Decision C(2013) 1573) that were completed and updated on 30 April 2015. The Commission services have organised closure seminars for Member States and have updated their risk assessments and consequently the Audit Strategy 2007-2013 to address any potential risk in view of the last part of the implementation period. The Commission has established a Task Force for Better Implementation of the Cohesion Policy funds in eight Member States. While emphasising full respect of legality and regularity of expenditure, this Task Force assisted the concerned Member States in the closure preparatory process. The same proactive and targeted approach will be applied to support the implementation of Member States’ programmes in 2014-2020, taking account of the lessons learnt. Commission has initiated further simplification measures in 2015 In July 2015, the Commission has set up a High Level Group of Independent Experts on Monitoring Simplification for Beneficiaries of the European Structural and Investment (ESI) funds. This group has been tasked to:

assess the uptake of simplification opportunities by Member States including Member States' commitments to reduce the administrative burden for beneficiaries as set out in their partnership agreements covering the 2014-2020 ESI funds;

analyse the implementation of simplification opportunities in Member States and regions, taking account of a study on the use of new provisions on simplification during the early phase of the 2014-2020 programme period, and their impact on administrative burden and costs; and

identify good practice in matters concerning the reduction of administrative burden on beneficiaries.

THE COURT’S OBSERVATIONS Audit scope and approach Annex 1.1 sets out the key elements of our audit approach and methodology. For the audit of Economic, social and territorial cohesion the following specific points should be noted:

(a) We examined a sample of 223 transactions as defined in Annex 1.1, paragraph 7. The sample is designed to be representative of the range of spending under the MFF heading. In 2015, the sample consisted of transactions from 15 Member States.

(b) For financial instruments under shared management, we also reviewed their disbursement rates (i.e. the proportion of funds used at the level of final recipients). This was done on the basis of the Commission’s progress reporting for 2014 and our examination of seven ERDF financial instruments in seven Member States.

(c) We assessed other elements of internal controls and annual activity reports:

pt_11

(i) the Commission’s supervisory activities of audit authorities in 16 Member States;

pt_12

(ii) the annual activity reports (AARs) included reports of Directorate-General for Regional and Urban Policy and of the Directorate-General for Employment, Social Affairs and Inclusion.

We checked, for a second year, whether and to what extent a performance measurement system was in place to assess whether the completed examined projects have achieved their output and result objectives as set out in the projects’ approval documents. We also verified if these objectives were in line with the OP objectives. This was done for the 149 of the 223 examined projects which were completed at the time of the audit. The results of this exercise are presented in the section on the assessment of projects’ performance. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES REGULARITY OF TRANSACTIONS Annex 6.1 contains a summary of the results of transaction testing. Of the 223 transactions examined, 72 (32 %) were affected by error. On the basis of the 33 errors we have quantified, the estimated level of error is 5,2 %. The Commission notes that the level of error reported by the Court is an annual estimate which takes into account corrections of project expenditure or reimbursements affected by errors detected and recorded before the Court’s audit. The Commission underlines that it is bound by the Financial Regulation, which stipulates, in Article 32(2)(e), that its internal control system should ensure, among other things, adequate management of the risks relating to the legality and regularity of the underlying transactions, taking into account the multiannual character of programmes as well as the nature of the payments concerned. The Commission will continue to exercise its supervisory role, in particular by implementing financial corrections and recoveries at the level that corresponds to the level of irregularities and deficiencies identified. The Commission further notes that given the multiannual character of the management and control systems under Cohesion policy, errors made in 2015 may also be corrected up to and at closure, as illustrated in section 2.4.3 of the respective 2015 AARs of DG Regional and Urban Policy and DG Employment, Social Affairs and Inclusion. The Commission notes that the estimated level of error for 2015 is in line with the error rates presented by the Court for the five last years. This confirms that the level of error for the 2007-2013 programming period remains stable and significantly below the rates reported for the 2000-2006 period. This development derives from an improved regulatory framework including simplification, strengthened management and control systems and the Commission’s strict policy to interrupt/suspend payments as soon as deficiencies are identified, as reported in the 2015 AARs. The Commission will continue to focus its actions on the most risky programmes/Member States and implement corrective measures when needed through a strict policy of interruptions and suspensions of payments up to closure, and to apply strict procedures at closure to exclude any remaining material risk of irregular expenditure. The Commission also notes that the error frequency considerably decreased compared to previous years. For the 2014-2020 period the Commission’s corrective capacity is further strengthened by providing for net financial corrections and therefore reducing the possibility for Member States to reuse funds, for significant deficiencies detected by the Commission. This will be an important incentive for Member States to detect and correct serious irregularities before certifying annual accounts to the Commission. Chapter 1 contains an assessment of the accuracy of the figures for financial corrections presented in the financial statements discussion and analysis of the consolidated EU accounts (see paragraphs 1.37 to 1.43). Chapter 1 of the 2012 annual report explained also how financial corrections are taken into account when calculating the estimated level of error. The Commission notes the Court’s willingness to take into consideration financial corrections applied in its calculation of the estimated level of errors. Such corrections stem from the Members States’ own corrective capacity in some cases — an important pillar of the assurance process — or derive from the Commission’s supervisory actions that may include imposition of flat-rate financial corrections to programmes where it considered that there are remaining serious deficiencies and risks. The Commission refers to its replies under paragraph 1.38. Figure 6.2 presents the extent to which the different types of errors contributed to our estimated level of error for 2015. Economic, social and territorial cohesion — Breakdown of the estimated level of error Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The main sources of error for Economic, social and territorial cohesion are the inclusion of ineligible expenditure in the beneficiaries’ cost declarations and the selection of the wholly ineligible projects followed by infringements of EU and national public procurement rules and state aid rules. Over the last four years, we did not detect any quantifiable error related to the use of simplified cost options (SCOs). For all funds during the 2014-2020 programming period, Member States are obliged to use SCOs for small projects under 50000 euro and can decide to apply SCOs more widely. The Commission considers that the Court’s assessment confirms that simplified cost options are less prone to errors and therefore has actively worked with programme authorities since the introduction of the simplified cost options (SCOs) to progressively extend their use. This has already led to positive results. The Commission continues to actively promote the use of SCOs in the 2014-2020 programming period, where they have been significantly strengthened both in the Common Provisions Regulation and in the ESF specific regulation, in order to reduce the administrative burden on the beneficiaries, promote result-orientation and further reduce the risk of error. The Commission intends to make proposals in 2016 to expand the use of SCOs for all ESI Funds in the Common Provision Regulation. Besides providing extensive guidance on the implementation of SCOs, DG Employment, Social Affairs and Inclusion is currently conducting a second round of simplification seminars in certain priority Member States which have experienced recurring high error rates in the 2007-2013 programming period and have not yet made sufficient use of SCOs. Ineligible expenditure When declaring costs to the Commission, national authorities certify that these costs have been incurred in compliance with a number of specific provisions laid down in EU regulations, national legislation or eligibility rules, specific OP rules, calls for interest, decisions approving projects for co-financing or grant agreements. Ineligible expenditure is the main source of error for both ERDF/CF and ESF. We found that ineligible costs had been declared in 8 % of the transactions examined (see Figure 6.3). These cases account for 55 % of all quantifiable errors and make up approximately 2,4 percentage points of the estimated level of error. The Commission will follow up the cases identified by the Court and propose actions as it deems necessary. The Commission notes that in some cases national or regional rules applied to Cohesion Policy expenditure are more demanding than those foreseen in the EU or the national legislation for similar nationally funded expenditure. These additional requirements can be seen as an instance of unnecessary administrative burden and complexity to Cohesion Policy expenditure imposed by the Member States to themselves and their beneficiaries, as described for example in the reports on gold plating and simplification issued by DG Employment, Social Affairs and Inclusion in 2013. The Commission also notes that three errors occurred because national rules were stricter or more complex than necessary (gold-plating). The Commission estimates that the breaches to these national rules contributed one percentage point to the error rate. Example of ineligible costs declared Ineligible costs declared: An ERDF project in United Kingdom provides financial support to SMEs for consulting services to promote and improve their products or services. However, only a part of the funding granted was given as a grant to SMEs and around 13 % was retained by the beneficiary. The amounts retained do not meet the eligibility conditions for co-financing since this expenditure was not incurred, not paid and not recorded as costs in the accounts of the beneficiary. As a result, a part of the grant is ineligible for co-financing. Ineligible projects We identified four ERDF projects for which the eligibility conditions set out in the regulations and/or the national eligibility rules were not met. These projects account for 12 % of all quantifiable errors making up approximately 1,5 percentage points of the estimated level of error (see Figure 6.4). The Commission will follow up the cases identified by the Court and propose actions as it deems necessary. Example of an ineligible project Ineligible beneficiary: For an ERDF project in the Czech Republic, the call for proposals specified that only SMEs were eligible. The managing authority provided funding to a beneficiary on this basis, which however had not been confirmed at the time the project was selected. Other ineligible projects were also found in the Czech Republic, in Italy and Poland. Breaches of public procurement rules Public procurement rules are a key instrument for spending public money economically and effectively and for establishing an internal market within the EU. Public procurement rules are applicable for all public spending in the Member States and are not specific to Cohesion policy. Non-compliance with EU or national public procurement rules has been over years a major source of errors in this policy area, in particular for regional and urban policy, mainly due to the types of projects co-financed. The Commission has therefore taken various preventive and corrective actions since the last programming periods in order to address weaknesses identified in that area. The Action Plan on public procurement set up in 2013 and politically endorsed by the Commission in December 2015 aims at further improving the implementation of public procurement rules in the Member States through additional preventive measures. The legal framework for ESI Funds 2014-2020 has also introduced ex ante conditionalities for the effective and efficient use of Union funds, which cover inter alia Member States’ public procurement systems. Suspensions of the interim payments may be decided in case of failure to fulfil an applicable ex ante conditionality by the end of 2016 (Article 19 of Regulation (EU) No 1303/2013). The Commission will follow up all errors reported by the Court in accordance with the Commission Decision C(2013)9527 final which sets out guidelines for determining financial corrections to be made by the Commission to expenditure financed by the Union under shared management, for non-compliance with the rules on public procurement. The 2014 Directives on public procurement which introduced simplifications will start producing their effect on the ground once transposed (see also Commission reply to paragraph 6.25). In 2015, we examined 140 public procurement procedures related to contracts for works and services underlying the expenditure for the transactions tested. The estimated contract value for these public procurements amounted to approximately 2,9 billion euro. The vast majority of these contracts are for projects co-financed by ERDF/CF OPs. We identified instances of non-compliance with EU and/or national public procurement rules for 26 of the 140 procedures examined. For four of the cases, a lower number than in previous years, the failures to comply with these rules were so serious that we could determine that the contract should have been awarded to a different bidder or that competition was restricted in such a way that led to an unjustified direct award. We classified these errors as quantifiable (see Figure 6.5). These errors account for 12 % of all quantifiable errors and make up approximately 0,7 percentage points of the estimated level of error. For estimating the level of error we used the provisions of the 2014 public procurement directive which were less stringent than those included in the 2004 public procurement directives (i.e. modifications of contracts). Example of serious failure to comply with public procurement rules Additional works awarded without proper public procurement: For an ERDF road construction project in Germany, additional works in excess of 50 % of the initial contract value were directly awarded to the same contractor. This is a breach of Article 31 of the public procurement directive 2004/18/EC. As a result, these costs are ineligible for EU co-financing. Similar cases were found in other ERDF projects in Italy and the United Kingdom. Other errors relating to tendering and contracting procedures occurred in a further 22 of the 140 public procurements examined. These errors include cases of the incorrect application of the selection criteria, shortcomings in the tender specifications and non-compliance with information and publication requirements. These errors do not contribute to the error rate we estimated. The Commission will follow up the cases identified by the Court and propose actions as it deems necessary. For two transactions in two Member States, the national authorities had identified serious cases of non-compliance with public procurement rules and imposed specific financial corrections at project level. These corrective measures have been taken into account in the calculation of our estimated level of error, since the corrective measures were taken before the notification of our audit. The Commission notes that adequate corrective action is indeed taken by some programme authorities in the quoted Member States and acknowledges that it is reflected by the Court in its quantification of errors. The objective of the Commission’s action plan referred to in the reply to paragraphs 6.19 to 6.21 is to increase the occurrence of such proactive corrective actions at Member State level. On 26 February 2014, the Council of the European Union and the European Parliament adopted three directives aimed at simplifying public procurement procedures and making them more flexible. EU countries had until April 2016 to transpose the new rules into national law (except with regard to e-procurement where the deadline is October 2018). In May 2016, the Commission has addressed letters of formal notice whereby the Commission reminds them of their overdue obligation to enact in national law the three EU directives to 20 out of the 28 Member States. If Member States continue to fail in their obligation to enact in national law, the Commission may open infringement procedures under Article 258 TFEU which could eventually lead to the imposition of fines to the Member States concerned. The Commission is closely monitoring the state of transposition in Member States. As of 30 June 2016, 16 Member States still have to transpose the directive on public procurement (2014/24/EU), while 19 have to transpose the directive on the award of concession contracts (2014/23/EU) and 17 the directive on procurement by entities operating in the water, energy, transport and postal services sectors (2014/25/EU). We have also published a special report on the actions taken by Member States and the Commission to address the problem of public procurement errors in the area of Economic, social and territorial cohesion. The Commission has accepted all recommendations made by the Court in its special report and is taking the recommended actions. The Commission refers to its replies to paragraphs 6.7 to 6.15 and 6.19 above. Infringements of state aid rules State aid is in principle incompatible with the internal market since it may distort trade between Member States. There are, however, some exceptions to this rule. The Commission directly enforces the EU state aid rules. Member States must notify all cases of state aid to the Commission (either through a scheme or case-by-case for a project), unless the project is below a de minimis ceiling or is covered by the General Block Exemption Regulation (GBER). For all cases notified, the Directorate-General for Competition forms a view on whether the aid is compatible with the internal market The ESF is less prone to infringements of state aid rules since very often projects fall under the de minimis rule. In case of a State aid notification, the Commission assesses whether the support constitutes State aid and, if so, whether it is compatible with the internal market. The outcome of the assessment leads to a Commission decision. We identified this year seven ERDF/CF projects in five Member States that infringed the EU state aid rules. Where necessary, we requested and obtained a preliminary assessment by the Directorate-General for Competition. This assessment and the case law of the European Court of Justice were taken into account when classifying errors. The Commission will follow up the cases identified by the Court and propose actions as it deems necessary. For three of these projects, we consider that the project should have obtained less or no public funding from the EU and/or the Member State according to the state aid rules. The main reasons for non-compliance with state aid rules refer to the excess of aid intensity limits These quantified errors make up approximately 0,3 percentage points of the estimated level of error (see Figure 6.6). We applied the 2014 GBER for estimating the level of error in this category. Example of a project infringing state aid rules The aid intensity limit was exceeded: For a CF project in Poland related to the construction of a port infrastructure, funding exceeded the maximum allowable under state aid rules for this type of project. Other cases were found in another ERDF projects in Latvia and the United Kingdom. We have also published a special report on the Commission’s and Member States’ efforts needed to prevent, detect and correct infringements of state aid rules in the area of Cohesion. The Commission refers to its replies to the Court’s observations and recommendations published in the referred special report and will follow up the accepted recommendations. Value-added tax declared by public bodies Value-added tax (VAT) is an EU indirect tax on the domestic consumption of goods and services levied at each stage in the chain of production and borne by the end consumer. According to EU legislation, recoverable VAT is not eligible for co-financing under ERDF/CF and ESF. Non-recoverable VAT can, thus, be declared for co-financing. The regulatory provisions mentioned by the Court for Cohesion policy are the outcome of a policy choice made by the co-legislators. They are overall aligned with the standard principles on eligibility of VAT for grants in other EU policies, in particular under direct management. When public bodies do not declare and therefore do not recover VAT, both the Commission and the Member States have interpreted the regulations as allowing VAT to be an eligible cost. However, the distinction between recoverable and non-recoverable VAT in the regulations is a source of confusion. In principle, the Commission relies on the Member State’s assessment of the beneficiary’s taxable or non-taxable status regarding VAT. However, as confirmed by the Court of Justice in 2012, the notion of recoverable VAT does not result only from the formal examination of the taxable/non-taxable status of the beneficiary. The Commission intends to provide clarification to Member States in that respect for the current 2014-2020 programming period. The Commission will follow up the cases identified by the Court. For example, the Court of Justice ruled in 2012 that VAT must be considered recoverable whenever an operator charges fees with VAT to the ultimate users of the infrastructure. We identified three cases where recoverable VAT was declared by public bodies as eligible cost (see Figure 6.7). Example of ineligible VAT Recoverable VAT is ineligible for EU co-financing: For an ERDF project in Hungary related to the rehabilitation of an unused mine to construct a new open-air event centre the beneficiary, a public municipality, declared VAT as eligible expenditure. However, the operator of the centre will charge fees with VAT to the ultimate users of the infrastructure. As a result, VAT expenditure is ineligible for EU co-financing. Similar cases were found in two ERDF projects in Germany. Even when VAT declared as expenditure by national governments acting as beneficiaries has been considered as eligible, the VAT as such does not represent a net cost to the Member State. A large infrastructure project can be implemented, for example, by a ministry or a body directly under its authority and include VAT in its costs since that ministry does not recover VAT paid. At the same time, the national budget receives the VAT that has been reimbursed by EU funds. The Commission notes that VAT is a general tax collected by national or regional governments which do not redirect it back to individual financed projects. A more stringent position which would declare VAT as ineligible costs for ministries or bodies directly under their authority would actually penalise those which are not in a position to recover VAT or be compensated for it. Such bodies would therefore be obliged to find additional funding to implement their projects or would have to contribute more to their financing than other beneficiaries recovering VAT. This is not the policy choice made by the co-legislators for cohesion policy. In the EU the standard VAT rates applied for most of the projects range from 17 % to 27 %. When VAT declared as non-recoverable by the ministries implementing the projects exceeds the national co-financing rate, the EU reimbursement may even exceed the actual costs incurred for the project, net of VAT. For example, this occurs when the co-financing rate of a project is 85 % or higher and the VAT rate paid by the ministry implementing the project is beyond 20 %. If the VAT rate paid would be 23 %, then the total EU contribution for the project would account for 104,6 % of the actual costs incurred net of VAT. This is particularly relevant for the Cohesion Fund where most of the projects have a co-financing rate of 85 % or higher. Insufficient reliability of checks at Member State level In 18 cases of quantifiable errors made by beneficiaries, national authorities had sufficient information to prevent or detect and correct the errors before declaring the expenditure to the Commission. If all this information had been used, the estimated level of error for this chapter would have been 2,4 percentage points lower. In addition, we found that for two cases, the error that we detected was made by managing authorities and intermediate bodies in Member States. These errors contributed 0,6 percentage points to the estimated level of error. The Commission is strictly following up these cases and agrees that sound and timely management verifications must be in place in order to prevent irregularities occurring in the first place or being included in payment claims. Since 2010, the Commission has been carrying out targeted audits on management verifications of high-risk programmes where it has identified that deficiencies could remain undetected or not detected in a timely manner by the programme audit authority. For DG Regional and Urban Policy and DG Employment, Social Affairs and Inclusion, results of these audits conducted up to 2015 are presented in their respective 2015 AARs (see pages 65 and 56/57, respectively). The Commission refers to the reinforced procedures in the regulatory framework for the 2014-2020 programming period, where management verifications and controls (including on-the-spot checks) have to be carried out on time for the certification to the Commission of programme accounts and submission of management declarations by the managing authorities on an annual basis. Audit authorities have to estimate reliable residual levels of error in the accounts as a result of all verifications, controls and corrections made since the end of the accounting year. The Commission considers that these reinforced control procedures should result in lasting reductions of the error rate. Furthermore, the Commission has developed new guidance in order to further strengthen the reliability of management verifications and to clarify the increased accountability of managing authorities and their intermediate bodies through management declarations and annual summaries in the 2014-2020 programming period. This guidance, which draws on the lessons learned from the previous programming period, was published in September 2015. THE COURT’S OBSERVATIONS EXAMINATION OF FINANCIAL INSTRUMENTS UNDER SHARED MANAGEMENT Financial instruments provide assistance to enterprises or urban projects by way of equity investments, loans or guarantees. They can be used in mainly three areas: for the support of small-medium enterprises (SMEs), for urban development and for the promotion of energy efficiency. By the end of 2014, 1025 financial instruments had been set up under 183 ERDF and ESF OPs in all but three Member States (Ireland, Croatia and Luxembourg). Taken together, they have an endowment of around 16018 million euro. In general, the funds implementing these financial instruments receive a contribution from the OP when their own legal structure is set up, and subsequently use this money to support projects. Such financial support can be provided only to projects which fall within the scope of the OP. Financial instruments are designed to have a revolving character or, for certain types of guarantee funds, to achieve a high leverage effect. Any resources returned from investments or loans made, including profits, are to be used again for the purpose of the activities implemented by the financial instrument. Only disbursements made to final recipients and fund managers’ costs and fees are considered eligible. Unused endowments of financial instruments are to be returned to the EU budget at closure. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Slow implementation of financial instruments The average disbursement rate for 1025 ERDF and ESF financial instruments was 57 % at the end of 2014. This represents a 10 percentage points increase in comparison to 2013 and a 20 percentage points increase in comparison to 2012. The Commission notes that the average disbursement rate of 57 % (at end of 2014) — an increased rate compared to 47 % at the end of 2013 — is an average at EU level that does not reflect the heterogeneous situation at programme level (see the Commission report published on 1 October 2015). The low disbursement rate at the end of 2014 is a result of a number of factors including the financial crisis, the limited experience in some Member States and the late start of some instruments (even in 2014). The Commission would further like to stress that taking into account the life cycle of the financial instruments, an increase is expected in the last year of implementation. Any conclusion on the effective disbursement rates can be fully drawn only at closure. In any event, at closure, initial amounts in financial instruments not disbursed to final recipients will be considered ineligible, recovered by the EU budget and lost for the programmes and Member States. The Commission believes that its report to be published on 1 October 2016 on the situation at end-2015 will show further progress in implementation. Figure 6.8 presents the disbursement rates for financial instruments financed by Economic, social and territorial cohesion as at 31 December 2014. Already in our last annual report we have pointed to the difficulties of several Member States to use the endowments of the financial instruments in full. We note particular problems with low disbursement rates in six Member States (Greece, Spain, Italy, Netherlands, Austria and Slovakia) compared to the EU average. As already mentioned in its reply to the Court’s Annual Report for year 2014 (paragraph 6.49), the Commission has undertaken together with the Member States a number of measures to ensure that the remaining investments made by financial instruments will reach final recipients in a timely manner. The Commission continued its efforts in this respect also in 2015. Disbursement rates for financial instruments financed by Economic, social and territorial cohesion as at 31 December 2014 Source: European Court of Auditors based on EGESIF_15-0027-00, 23 September 2015. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Information on implementation of ERDF and ESF financial instruments is based on data received directly from the Member States. Since 2011, the Commission has made substantial efforts to improve the quality of these data. Nevertheless, the data presented in the Commission’s report are in some cases incomplete or inaccurate. This was also the case for four of the seven financial instruments that we examined in 2015. In May 2015, the Commission provided guidance on reporting instructions to Member States, clarifying the reporting requirements. Furthermore, the Commission has improved its reporting tools for 2014-2020: Member States will have to report the data required under Article 46 of the Common Provisions Regulation which is more comprehensive than the data required for 2007-2013. In April 2016, the Commission services presented a further guidance to Member States on this reporting. Ineligible final recipients In two of the seven financial instruments examined guarantees or loans were approved for final recipients who were ineligible according to the regulation and the national law. These cases are classified as quantifiable errors either under the ineligible expenditure or under the ineligible projects category (see Figure 6.9). The Commission notes that for two errors linked to ineligible recipients in financial instruments, it cannot at that stage assess the impact of these individual errors on the EU budget. As indicated at point 3 of the Guidelines for determining financial corrections to be made to expenditure co-financed by the EU under the structural funds and the European fisheries fund for non-compliance with the rules applicable to financial engineering instruments for the 2007-2013 programming period, individual irregularities can be replaced with legal and regular investments carried out by the financial instruments in final recipients by partial or final closure. The Commission therefore considers that, without contesting the underlying errors, it cannot assess the impact of such errors on the EU budget at this stage. Example of an ineligible loan (ineligible recipient) Ineligible final recipients: In the case of one of the ERDF financial instruments in Hungary, the financial intermediary issued a loan for a company belonging to its own group. However, the regulation and the applicable Commission guidance explicitly forbid that a company belonging to the same group can be considered as an eligible final recipient. As a result the loan is ineligible for co-financing. Other similar cases of ineligible recipients were found in Italy. Extension of the 2007-2013 eligibility period through a Commission decision only We recall a specific issue that was disclosed in the 2014 annual report. Article 56(1) of Regulation (EC) No 1083/2006 established the 31 December 2015 as the end of the eligibility period of payments. During 2015, the Commission decided to extend the eligibility period until 31 March 2017 through a Commission Decision instead of asking the Council and Parliament to amend the Regulation. Since there is still no legislation of sufficient hierarchy we will consider disbursements made after 31 December 2015 to be irregular. The Commission expressed its position on the same observation already in its reply to the Court’s Annual Report for the year 2014. The Commission has acted in line with the European Council’s recommendations of December 2014 and within the margin offered by the existing regulatory framework. Therefore, the Commission considers the disbursements made to final recipients made until end March 2017 as eligible expenditure. The modification decision on the closure guidelines of the Commission does not affect Article 56(1), according to which the contribution from the funds into financial instruments must have been paid at latest at 31 December 2015. This remains the rule. The modification clarifies the Article 78(6) of Regulation (EC) No 1083/2006. Article 78(6) of Regulation (EC) No 1083/2006 is a lex specialis to Article 56(1) of that Regulation. On this basis the Commission concludes that:

Article 78(6) implies that the contribution from the operational programme to the financial engineering instrument has to be paid until 31 December 2015,

as regards eligibility of expenditure paid by a financial engineering instrument to the final recipient, Article 78(6) allows such expenditure to be eligible until the date of submission of closure documents,

as 31 March 2017 is the final date for submission to the Commission of the closure documents, the modification of the closure guidelines clarifies that closure as referred to in Article 78(6) of the General Regulation is the date of submission of the closure documents, i.e. 31 March 2017 (see also Commission reply to paragraph 6.52 and to Recommendation 5 of the Court’s 2014 annual report).

The Commission therefore disagrees that disbursements made after 31 December 2015 are to be considered irregular. We have also published a special report on the lessons to be learnt from the implementation of the EU budget through financial Instruments for the 2007-2013 programming period. The Commission refers to its replies to the Court’s recommendations published in the referred special report and is already taking actions to follow up the accepted recommendations. EXAMINATION OF ELEMENTS OF INTERNAL CONTROL SYSTEMS, INCLUDING ANNUAL ACTIVITY REPORTS Assessment of the Commission’s supervision of audit authorities Commission uses the work of audit authorities in Member States to estimate the level of error Audit authorities provide assurance to the Commission as to the effective functioning of the management systems and internal controls for an OP and the legality and regularity of the expenditure certified. This information is provided by audit authorities in their annual control reports, audit opinions and system audit reports. Since 2009, the Commission has carried out enquiries of the main audit authorities, including on-the-spot reviews of their work, to assess the level of reliance they can place on them. The Commission is closely cooperating and coordinating with audit authorities, and started reviewing their methodologies and audit results as early as 2009. This contributed to capacity building by providing advice, guidance and recommendations to audit authorities through the Commission’s re-performance of audit work carried out by audit authorities. The Commission underlines that its assessment of the reliability of the work of audit authorities is to be seen in that context. As a result of its audit programme that entailed 316 audit missions carried out on the spot cumulatively since 2009 covering 51 audit authorities in charge of auditing around 98,5 % of ERDF/CF allocations for the 2007-2013 period, DG Regional and Urban policy concluded by end 2015 that the work of the reviewed audit authorities can in general be relied upon. Concerning DG Employment, Social Affairs and Inclusion, at the end of 2015, 236 audit missions have been carried out since 2009 covering 89 audit authorities out of 92 (96,7 %) in charge of 115 out of 118 OPs representing 99,6 % of the expenditure. The Commission concluded that it can rely on 82 audit authorities covering 98,2 % of ESF allocations. In their 2015 AARs, DG Regional and Urban Policy and DG Employment, Social Affairs and Inclusion provided a detailed assessment of the accuracy and reliability of the audit information and results reported by audit authorities in their 2015 annual control reports in relation to 2014 expenditure (see pages 55 to 60 of DG Regional and Urban Policy’s 2015 AAR and pages 52 to 54 of DG Employment, Social Affairs and Inclusion’s 2015 AAR). The Directorates-General for Regional and Urban Policy and for Employment, Social Affairs and Inclusion make use of this information when preparing their annual activity reports (AARs) and throughout the year to decide on possible interruptions and/or suspensions of payments to OPs. In 2015, DG Regional and Urban Policy indicated in its AAR that approximately two thirds of the interruptions and pre-suspensions of payments were based on audit results reported to the Commission by audit authorities during the year or at year end (see page 68 of DG Regional and Urban Policy’s 2015 AAR). In the course of 2015, interruptions and pre-suspension procedures concerned 117 ERDF/CF programmes and almost EUR 9,1 billion of payment claims submitted by the Member States but not paid unless the Commission has obtained additional evidence that all necessary corrections had been made by the Member State concerned. For ESF, as reported on page 59 of DG Employment, Social Affairs and Inclusion’s AAR in 2015, the Commission sent 2 warning letters and 17 pre-suspension letters; it decided 27 interruptions of payments and suspended 10 operational programmes. In total, EUR 1,4 billion of payment claims were interrupted. In order to assess whether they can rely on this information, the directorates-general check the error rates reported by the audit authorities for each OP (or group of OPs):

if the Commission considers the error rate to be reliable (and representative for the expenditure certified), it accepts the rate reported by the audit authority. Also the Commission may recalculate the rate for its own assessment based on additional information obtained from the audit authority;

in the case of unreliable error rates, the Commission applies a flat error rate (between 2 % and 25 %) in line with the results of its assessment of the functioning of management and internal control systems.

The Commission underlines that this assessment of the reliability of error rates each year is based on a thorough desk analysis of all available information completed by risk-based on-the-spot fact-finding missions, also taking account of the overall assessment of the reliability of the work of audit authorities as a result of the comprehensive audit work referred to in the reply to paragraph 6.47. When necessary, the Commission requests and obtains from the audit authorities any additional information required. The Commission also calculates a cumulative residual risk for each OP (or group of OPs), which takes into account all financial corrections since the start of the programming period. This includes corrections already implemented at EU and/or national level. Based on these two indicators and additional information at its disposal, the Commission forms its assessment of the management and control system for the OP. This assessment is then reported in the annual activity report of the directorate-general. The Commission assesses the effective functioning of the management and control system for each programme and for each authority (managing, certifying and audit authorities), based on all available EU and national system audit results on 15 key regulatory requirements according to a methodology shared with audit authorities. All programmes are therefore assessed against audit opinions at national and Commission level based on audits carried out on systems and representative samples of operations. Error rates reported by Member States and residual error rates calculated by the Commission are important elements of this assessment, but not the only ones. Operational line managers and authorising officers by sub-delegation also assess the level of management assurance for each programme. For the purpose of the assurance disclosed in the AARs, the opinion for each operational programme is, therefore, the result of an overall assessment based on all information available to the Commission. This provides a comprehensive and detailed assessment of the effectiveness of management and control systems per Member State and, for operational programmes included in the 2015 reservations, information on the main issues identified and the actions taken. Commission’s assessment of annual control reports can only partially address the risk of unreliable information being presented by national authorities The robustness of the Commission’s assessment of ERDF/CF and ESF OPs depends on the accuracy and reliability of information reported by Member State authorities. Our audits in previous years identified two main risks in this regard:

audit authorities may understate the errors and/or may not extrapolate them properly. As a consequence, the reported error rates may not always be reliable;

information on financial corrections reported by Member States may not always be reliable or accurate and, as a result, the Commission’s calculation method may result in an understated cumulative residual risk.

As explained in their respective AARs, both directorates-general have taken measures to improve the reliability of audit information (error rates and audit opinions) and financial corrections reported by the certifying authorities. To mitigate the risks mentioned by the Court, the Commission is implementing two audit enquiries:

the Commission reliance on the audit information, reported by audit authorities, including the annual control reports, the audit opinions and error rates, is based on wide desk and on-the-spot audit work. This aims at verifying whether audit findings are confirmed, including through on-the-spot re-performance by Commission auditors. The desk review of annual control reports covers a wide range of verifications linked e.g. to the accuracy of the calculations, parameters used, treatment of errors detected or reconciliation of the audited population and is complemented by fact-finding missions (see also the Commission’s reply in paragraph 1.36 and 6.47). Additional specific information is requested and obtained from audit authorities when necessary in case of doubts but is not systematically requested in all cases, in line with the single audit principle and the regulatory requirements. As a result of this in-depth audit work, Directorates-General for Regional and Urban Policy and Employment, Social Affairs and Inclusion have assessed 95 % of the Member States error rates for their policy areas as a reliable source of information for the purpose of calculating the risk to 2015 payments (pages 58 and 53 of respective 2015 AARs). For the remaining 5 % where the error rates reported by the Member States were considered unreliable the Commission replaced them by a best estimate on the basis of flat rates,

specific risk-based on-the-spot audits (so far covering 75 OPs over the last three years) in order to ensure that the corrections reported are effectively implemented and, in case of doubts or insufficient evidence, deduction of the amounts concerned from the cumulative financial corrections taken into account for the purposes of the calculation of the residual error rate. As a result of their audits and desk consistency checks on the reliability of the article 20 statements on financial corrections submitted for all programmes, the Commission services decided to take a prudent approach and exclude some of the reported financial corrections for the calculation of the cumulative residual risk (see the annual activity report of Directorate-General Regional and Urban Policy, pages 72 and 73).

Contrary to 2007-2013, in the 2014-2020 programming period audit authorities will issue their audit opinion based on residual error rates after corrections have been implemented. As a result, the accuracy of financial corrections will be systematically audited as part of the audit on accounts. The Commission’s verifications of the annual control reports consist mainly of desk reviews and can therefore only partly address these risks. In particular, we consider that the Commission’s scope for validating (and, where necessary, adjusting) the reported error rates is limited when audit authorities are not requested to provide the Commission with more specific information on their audits of operations (for example details on the audit scope, coverage/sub-sampling and error classification) for its verification of the annual control reports. Our analysis showed that in 2015 such information has been requested more often than in previous years. During 2015, as in previous years, both directorates-general also performed on-the-spot checks on financial corrections implemented by ten Member States. For seven of these, the Commission identified deficiencies requiring adjustments to the financial corrections reported. In 2014, the Commission detected deficiencies for seven of twelve Member States visited. These findings confirm our own assessment and highlight, in the context of the new 2014-2020 programming period, the importance of ensuring that Member States put in place reliable reporting systems. Commission considers that expenditure declared for 58 % of all OPs was free of a material level of error For 2015, the Commission considers that it had assurance that 254 of the 440 ERDF/CF and ESF OPs (58 %) were free from a material level of error: audit authorities had reported error rates below the Commission’s materiality threshold of 2 %, and these rates had been validated by the Commission. These OPs represent around 38 % of the 2015 payments. In 2015, as in previous years, we examined whether the Commission was effective in verifying the accuracy and reliability of the error rates reported by audit authorities. This was carried out for a sample of 102 OPs in 16 Member States (80 of the 322 ERDF/CF OPs and 22 of the 118 ESF OPs) on the basis of the Commission’s working files. Our sample included OPs for which the Commission had accepted the rates reported by audit authorities, but also OPs for which rates had been recalculated or where flat rates had been imposed. Commission’s validation (or recalculation) of the error rates is consistent with the evidence provided by audit authorities In relation to five OPs covered by five ACRs, we identified weaknesses that had not been detected or properly addressed by the Commission. However, we consider that none of these cases should have led the Commission to make additional reservations (or reservations with a higher financial impact) in the 2015 annual activity reports. This is so since almost all of these OPs (or group of OPs) are already under reservation for different reasons. The Commission notes the improvement in the assessment of its verification of errors rates reported by the audit authorities. Improvements in audit authorities’ checks on state aid and public procurement As in previous years, we also examined whether national audit authorities carried out appropriate checks of public procurement and state aid rules. We noticed a significant improvement in that area compared to the previous year’s assessment since most of the check lists used by audit authorities in their audit of operations adequately covered compliance with state aid and public procurement rules. As a result, for two of the 102 OPs sampled the state aid checks carried out by the audit authorities were inadequate (for example providers of services of general economic interest were not covered or for notified aid schemes and research and development measures the aid intensity was not verified). For one of these two OPs the public procurement checklist is considered insufficient. Through its own verifications, the Commission also noted the significant improvement reported by the Court. During the ACR review, the Commission examines whether the audit authority’s opinion is solidly grounded on the basis of the results of system and operations audits. The examination of the work of the audit authorities with regards to state aid and public procurement is included, amongst other issues, in the Commission’s review on audit authorities which include an examination of the methodology for system audits and audits of operations (including checklists) as well as re-performing audits already done by audit authorities (see reply to paragraph 6.47). In some cases the Commission had already recommended to the audit authorities concerned, through its own audit reporting, to reinforce their checklists to better cover the verification of state aid and public procurement issues. The Commission will continue to ensure that audit authorities use appropriate checklists. Commission guidance requires audit authorities to include contributions to financial instruments to the audited population rather than checking the actual disbursements to final recipients Financial contributions from the OPs to funds implementing financial instruments for the whole period are generally made through a single payment as soon as the legal structure of the funds has been set up. Alternative financing arrangements can be specified in the funding agreements. In accordance with regulation, these payments are certified to the Commission as expenditure incurred. In line with the Commission’s guidance to audit authorities on sampling, ex post verification of such payment by audit authorities generally relate to the financial year in which the payment has been made. These checks generally have a limited scope because within the first year few operations are selected for support and limited amounts are disbursed to final recipients and projects. In line with Article 78(6) of Regulation (EC) No 1083/2006, the funds’ contribution into financial instruments is included in payment claims to the Commission. These payments constitute part of the population to be audited by audit authorities through sample checks in line with the updated Commission guidance. The audit framework developed in 2011 for ERDF/CF and ESF clearly foresees that audits should cover the set-up and the implementation of the financial instrument and that specific verifications by closure are envisaged. Audit authorities and Commission may carry out thematic audits on financial instruments which cover both the set-up of the fund and the actual implementation of projects through the audit of a sample of projects carried out by final recipients. By letter of 18 March 2015, the Commission has also clarified to the audit authorities that financial instruments should always be included in the population subject to sampling and has informed them that it considered good practice that these operations are audited as a separate stratum for the purposes of sampling. In addition, during the Homologues Group Meeting of September 2015, the Commission discussed with audit authorities the audit approach and audit strategy towards closure. The necessary assurance on financial instruments at closure will be obtained by combining results from system audits carried out during the implementation, audits of a sample of transactions which covered the implementation when this was possible and additional testing by closure to ensure that actual disbursements to final recipients are legal and regular. This discussion was systematically followed up in early 2016 in bilateral audit coordination meetings with the concerned Member State’s audit authorities. Finally, for the purpose of the closure declaration, the audit authorities will have to obtain and provide assurance on the legality and regularity not only of the OP contribution paid into financial instruments but also of the use of this contribution through underlying transactions implemented by the final recipients as listed in Article 78.6 of Regulation (EC) No 1083/2006. For three OPs the contribution to financial instruments accounts for a considerable part of the the population and of the audited sample. The audit authorities found these contributions to be free of error. Unless specific verifications are envisaged at closure, the actual implementation of financial instrument operations is unlikely to be checked by audit authorities before closure. In the last two years, we detected the selection of ineligible recipients in some of the instruments audited (see also paragraph 6.44). The Commission has not issued a formal horizontal guidance on the audit of the disbursements made from the financial instrument to final recipients up to closure. However, the audit approach is agreed upon bilaterally between the Commission and the Member State concerned. Where the financial instruments are managed by the European Investment Bank (EIB), audit authorities cannot check the contributions to the fund due to limitations imposed to access rights by EIB. Further discussions between the Commission and the EIB Group should allow managing and audit authorities to have assurance at closure on this specific point. Very low error rates reported by the audit authorities for OPs for which we have found errors For 110 of the 440 ERDF/CF and ESF OPs, the Commission has validated error rates below or equal to 0,5 %. For 16 of these OPs, we performed at least one examination during the last two financial exercises. For nine OPs we reported quantifiable errors in the projects examined (see paragraphs 6.55 and 6.56). The Commission notes that the fact that individual errors are found in a programme does not prejudge audit results based on statistical samples adequately and randomly selected or samples covering a significant part of the expenditure declared. The Commission considers that individual errors reported by the Court do not put in doubt the reliability of the reported and validated error rates in 2015. Examination of the Commission’s annual activity reports We assessed the 2015 annual activity reports (AARs) and accompanying declarations of the Directorates-General for Regional and Urban Policy, and Employment Social Affairs and Inclusion. In particular:

we checked the consistency and accuracy of the Commission’s calculation of the amounts at risk;

we assessed the reservations made for 2015.

Commission’s calculation of the amounts at risk for ERDF/CF and ESF OPs Overall, the Commission’s calculation of the amounts at risk in the 2015 annual activity reports is accurate and consistent with the available information reported and/or provided by the audit authorities. We recall however that the annual error rates reported by the Commission in the annual activity reports are not directly comparable to those estimated by us. The Commission agrees that the Court’s error rate and that of the Commission are not directly comparable. However, the objective of this process is essentially the same, i.e. assessment of the risk to the EU budget in a particular year. The Commission takes into account all the differences quoted by the Court in its special report 16/2013 (see Commission reply to paragraph 11 of this report). Except for these differences, the Commission considers that for the 2015 annual report, as was the case for the last five years in a row for Directorate-General for Employment, Social Affairs and Inclusion and four years in a row for Directorate-General for Regional and Urban policy before this annual report, the result of the Commission’s assessment is in line with the error rates calculated by the Court (see pages 65 and 100 of the respective AARs).

the Directorate-General for Regional and Urban Policy estimated that between 3,0 % and 5,6 % of the interim and final payments for ERDF/CF OPs of the 2007-2013 programming period authorised during the year were at risk of error. 63 % of payments made by the Commission were to OPs with a validated error rate above 2 %;

Directorate-General for Regional and Urban policy made in 2015 63 % of payments to programmes it assessed at year-end as affected by a material level of error above 2 %, but for the vast majority of these programmes sufficient corrections were made in order to bring the cumulative residual risk below materiality. See also Commission reply to paragraph 6.67.

the Directorate-General for Employment, Social Affairs and Inclusion estimated that between 3,0 % and 3,6 % of the interim and final payments for ESF OPs of the 2007-2013 programming period authorised during the year were at risk of error. 56 % of payments made by the Commission were to OPs with a validated error rate above 2 %.

Directorate-General for Employment, Social Affairs and Inclusion made in 2015 56 % of payments to programmes it assessed at year-end as affected by a material level of error above 2 %, but for the vast majority of these programmes sufficient corrections were made in order to bring the cumulative residual risk below materiality. See also Commission reply to paragraph 6.67. Commission’s reservations for ERDF/CF and ESF OPs In cohesion policy, the Commission issues reservations on OPs (or groups of OPs) when:

significant weaknesses in the OPs management and control systems have been identified, or

the OPs validated error rate exceeds 5 %.

For OPs with a validated error rate between 2 % and 5 %, the Commission also considers whether the OPs cumulative residual risk exceeds 2 %. The Commission notes that as reported in the respective 2015 AARs:

93 % of ERDF/CF programmes presented a cumulative residual risk below 2 % at the end of 2015 (page 58). In relation to the 21 programmes with a cumulative residual risk above 2 %, reservations were made for 20 programmes and for the remaining programme the exception was disclosed in the AAR (page 180),

90,7 % of the ESF programmes, representing 93,3 % of the 2015 payments, presented a CRR cumulative residual risk below 2 % (page 61). Reservations were made for the remaining programmes, unless the necessary corrective measures had already been implemented.

The number of OPs subject to reservation decreased in 2015 from 77 to 67 for ERDF/CF OPs and from 36 to 23 OPs for ESF. Meanwhile, the estimated financial impact of these reservations decreased from 234 million euro in 2014 to 231 million euro in 2015 for ERDF/CF and from 169,4 million euro in 2014 to 50,3 million euro in 2015 for ESF. The Court’s observation demonstrates the maturity of the methodology used for reservations in the respective AARs and the strict implementation of this methodology by DG Regional and Urban Policy and DG Employment, Social Affairs, and Inclusion. Both directorates-general follow a strict approach with regard to the interruption and suspension of payments. They also apply a preventive approach as regards payments, i.e. no payments are made to operational programmes for which an audit mission is planned until the audit mission is completed. Payments are resumed afterwards unless significant deficiencies were identified. We found that the reservations made by the Commission in the 2015 annual activity reports were in line with the Commission’s instructions and the information made available to the two directorates-general. The Commission estimates the cumulative residual risk to be 1 % of the payments for all ERDF/CF OPs and 0,9 % for all ESF OPs. These figures are based on the reported error rates by audit authorities and validated by the Commission and take also into account the financial corrections reported by certifying authorities to the Commission. Looking forward: 2014-2020 programming period In our 2014 annual report we signalled delays in the start of the 2014-2020 programming period. We note that at the end of 2015, fewer than 20 % of the national authorities responsible for ESI funds have been designated. While Member State authorities may for example launch calls and select projects, they cannot submit statements of expenditure to the Commission until their designation. As a consequence, there is a risk that delays in the budgetary execution for the 2014-2020 programming period will be greater than those for the 2007-2013 period. The Commission notes that designation is a Member States’ responsibility. The Commission has encouraged Member States on several occasions (in meetings and written communication) to speed up this process — while ensuring full respect of the designation criteria defined in the regulation — and provided clarifications and assistance where necessary. The Commission further notes that implementation on the ground can start well before the designation process is finalised at the level of the Member States and notified to the Commission. On average already around 10 % of the allocations have been committed to selected operations, with variations between Member States. CONCLUSION AND RECOMMENDATIONS The conclusion for 2015 Overall audit evidence indicates that spending on Economic, social and territorial cohesion is affected by a material level of error. The Commission notes that the estimated level of error presented by the Court remains stable and is also consistent with the error rates reported in the respective AARs, subject to the differences set out in paragraph 6.66. The Commission will continue to focus its audits and actions on the most risky programmes/Member States and implement corrective measures when needed through a strict policy of interruptions and suspensions of payments up to closure, and to apply strict procedures at closure to exclude any remaining material risk of irregular expenditure. The Commission further notes that given the multiannual character of the management and control systems under Cohesion policy, errors made in 2015 may also be corrected in subsequent years and before closure as illustrated in section 2.4.3. of the respective 2015 AARs of DG Regional and Urban policy and DG Employment, Social Affairs and Inclusion. For the 2014-2020 period, reinforced requirements on legality and regularity include the introduction of annual accounts, reinforced management verifications and accountability as well as audit opinions based on residual levels of errors each year, as well as a retention of 10 % on all Commission interim payment until all controls are performed at Member State level and annual accounts submitted. Residual level of errors will be calculated for each accounting year covering the accounts. The Commission’s corrective capacity is further strengthened by the possibility to apply net financial corrections for serious deficiencies under certain conditions, thereby reducing the possibility for Member States to reuse funds. This will be an important incentive for Member States to detect, to report and to correct serious deficiencies before certifying annual accounts to the Commission. All these elements should contribute to a lasting reduction in the error rate. See also Commission reply to paragraph 6.11. For this MFF heading area, testing of transactions indicates that the estimated level of error present in the population is 5,2 % (see Annex 6.1). Our audits since 2009 have shown that the level of error for the 2007-2013 programming period is significantly lower than for the 2000-2006 programming period. It remains, however, persistently above the materiality threshold. We also note that expenditure in this area is reimbursed based on declarations of actual costs incurred by beneficiaries. Several attempts at simplification have been made in previous years, but had only a limited impact on the regularity of transactions. At the same time, increased levels of control have caused a significant administrative burden on national administrations and beneficiaries. In 2015, the Commission set up a high level group of simplification in view of preparing for the post-2020 programming period (see paragraph 6.8). The level of error for the 2007-2013 programming period remains stable and significantly below the rates for the 2000-2006 period. This development derives from an improved regulatory framework including simplification, strengthened management and control systems and the Commission’s strict policy to interrupt/suspend payments as soon as deficiencies are identified, as reported in the 2015 AARs. Moreover, a study commissioned by the Commission in 2010 on the administrative costs of the implementation and management of the funds shows that their share of the total eligible ERDF/CF expenditure is not excessive as it was estimated to 3,2 % for the entire period (approximately EUR 12,5 billion). A further study in 2012 which was used as the basis for the impact assessment for the 2014-2020 period estimated that Commission proposals would lead to a reduction of almost 20 % in administrative burden for beneficiaries. The Commission has also been actively working with the Member States on the promotion of the simplified cost options, which are focused on results and based on calculated costs as opposed to reimbursement of actual expenditure incurred in order to progressively extend their use and considers that these efforts have already led to positive results (see also Commission replies to paragraph 6.15). In order to better understand the barriers to use of simplification measures, the Commission is undertaking a study which should enable it to assess the uptake of simplified cost options on Member State level and the potential for their increased use. The Commission has set up the High-level Group on simplification in order to make recommendations to maximise the possibilities for reducing administrative burden under the 2014-2020 regulatory framework. Recommendations Annex 6.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, we presented 16 recommendations. Of these recommendations, the Commission fully implemented nine recommendations, four were implemented in most respects, two were implemented in some respects, and one was not implemented. The Commission notes that it had not accepted the recommendation reported under this paragraph as not implemented. For further details, see Commission reply to Annex 6.2. Following this review and the findings and conclusions for 2015, we recommend that the Commission:

Recommendation 1: should fundamentally reconsider the design and delivery mechanism for the ESI funds when making its legislative proposal for the next programming period taking also into account the suggestions of the high level simplification group (target implementation date: final results before the Commission presents its proposal for the post-2020 period);

The Commission partially accepts the recommendation. While the Commission is not in a position at this stage to make any specific commitment in relation to legislative proposals for the post-2020 period, it will consider all viable options taking into account the suggestions of the High-Level Group as well as results of studies and evidence from 2014-2020 programme implementation and other relevant information.

Recommendation 2: should make use of the experience acquired in the 2007-2013 programming period and report on a focused analysis of the national eligibility rules for the 2014-2020 programming period. Based on the above, it should provide guidance to Member States on how to simplify and avoid unnecessarily complex and/or burdensome rules that do not add value with respect to the results to be achieved by the policy (gold-plating), in particular when they result in significant and/or recurring errors (target implementation date: end of 2018);

The Commission accepts this recommendation. It will pay specific attention to national eligibility rules in its monitoring and audit of national management and control systems, in particular in Member States with persistently high error rates On this basis, it will help the Member States concerned to simplify and avoid unnecessarily complex and burdensome rules, facilitate the exchange of good practices and will report on the experience acquired in 2018. The Commission continues to heavily invest on the use of simplified cost options during the 2014-2020 programming period. The Commission recalls that, under shared management, it is primarily the responsibility of the Member States to review and simplify their national eligibility rules.

Recommendation 3: should submit a legislative proposal to amend, through a legislative act of equal legal value, Regulation (EC) No 1083/2006 with respect to the extension of the eligibility period for financial instruments under shared management to the Council and the Parliament. A change of the regulation is required to provide legal certainty to the Member States (target implementation date: immediately);

The Commission does not accept the recommendation. As it has stated in its reply to the same recommendation in the 2014 Annual report of the Court of Auditors, it considers that, within the mandate given by the European Council in December 2014, the modifications introduced in its closure guidelines were within the scope of Article 78(6) of Regulation (EC) No 1083/2006, as amended, and therefore did not require an amendment of the legislative act (see also Commission common reply to paragraph 6.59 to 6.62 above and to Recommendation 5 of the Court’s 2014 annual report). The Commission notes that the Council and the European Parliament took note of, but did not support, the recommendation made by the Court in the context of the 2014 discharge.

Recommendation 4: for the 2014-2020 programming period, the Commission should clarify to Member States the notion of recoverable VAT, in particular for public beneficiaries, to avoid different interpretation of the term non-recoverable VAT and avoid a sub-optimal use of EU funds (target implementation date: mid 2017);

The Commission accepts this recommendation. It will provide guidance to Member States to clarify the notion of recoverable VAT in the context of the 2014-2020 programming period, within the policy choice of the co-legislators for value added tax (VAT) eligibility, as provided for in Article 69(3)(c) CPR.

Recommendation 5: should ensure that all the expenditure related to ERDF and ESF financial instruments for the 2007-2013 programming period are included sufficiently early in the closure declarations to enable audit authorities to carry out their checks. In addition, the Commission should encourage all Member States that implemented financial instruments to carry out specific audits on the implementation of these instruments in view of the closure (target implementation date: immediately).

The Commission accepts the recommendation. It already clarified in its closure guidelines for 2007-2013 that in order for the audit authorities to have sufficient time to carry out their work for the closure declaration, the application for payment of the final balance and the final statement of expenditure should be submitted to the audit authorities well in advance, at a date to be agreed between authorities for each programme. For the purpose of the closure declaration, the Commission has already encouraged audit authorities to obtain and provide assurance on the legality and regularity not only of the programme contribution paid into financial instruments but also of the use of this contribution through underlying transactions implemented by the final recipients. The Commission will on the basis of the closure packages submitted by 31 March 2017 assess if sufficient audit work was carried out on financial instruments as the eligibility of underlying expenditure is determined at closure for such instruments. THE COURT’S OBSERVATIONS ASSESSMENT OF PROJECTS’ PERFORMANCE Implementing the EU budget under the principle of sound financial management requires that spending should not only focus on complying with the rules but also on the achievement of intended objectives. 149 of the 216 projects we examined were completed at the time of the audit. All 15 Member States sampled had at least three completed projects that could be assessed (see paragraph 6.9). In addition to checking the regularity of these 149 transactions, we assessed for a second year, based on information provided by the beneficiaries, whether and to what extent:

a performance measurement system was in place to assess whether the objectives specified for projects in the project’s approval documents (project application, grant agreement, contract and/or decision for co-financing) were in line with those set out in the OPs and whether appropriate result indicators were set in order to measure the projects’ performance;

projects had achieved those objectives and met the targets set for each indicator. Our analysis was broader for this year compared to the previous year as we did not focus only on project outputs, but also on results in order to measure the projects’ performance in contributing to the OP result objectives.

Figure 6.10 summarises the results on the assessment of the completed projects’ performance. The results on the assessment of the completed projects’ performance Performance assessment No of projects for which: Both output and result indicators were defined and assessed Outputs were defined and assessed but result indicators were not defined Neither outputs nor result indicators were defined and could not be assessed TOTALS Number of projects 90 57 2 149

Fully achieved

52 (58 %) 50 (88 %) — 102 (68 %)

Partially achieved

34 (38 %) 7 (12 %) — 41 (28 %)

Not achieved

4 (4 %) — — 4 (3 %)

Cannot be assessed

— — 2 (100 %) 2 (1 %) THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES For all but two of the projects we examined there was a performance measurement system in place linking the project to the OP objectives For 147 of the 149 projects (99 %) we conclude that there was a performance measurement system in place based on which an assessment of whether the output indicators are in line with the operational programme indicators can be made (as specified for each priority axis). For the remaining two cases (1 %), we could not assess the output of the projects either because specific targets in the output objectives were not set at the stage of the approval of the project or the managing authority/intermediate body could not provide appropriate documentation to us to substantiate that such targets had been specified and communicated to the beneficiary. Therefore, we cannot assess whether or to what extent these projects contributed to the achievement of the OP objectives. Almost all of the 90 projects with both output and result objectives defined had fully or partially met them For the 2007-2013 programming period there was a legal requirement obliging beneficiaries to define and report on outputs but not on results. Performance had to be monitored at the level of the priority axes of the OP and not at project level. Although the structural funds legislative framework for 2007-2013 did not oblige national authorities to define result indicators at project level, the Commission encouraged this whenever relevant. Some result indicators could indeed not be meaningfully measured at the level of a single project. However, the Member State authorities still had the possibility to define result indicators. This allows linking the projects’ results to the priority axis and determining the contribution of the projects to the achievement of the objectives of the operational programme. Result indicators were defined for 90 of the 149 (60 %) completed ERDF/CF and ESF projects. 52 of the 90 completed projects (58 %) which could be assessed at both output and result objectives level had fully achieved both the output and the result objectives set by the managing authorities (see Figure 6.11). Another 34 (38 %) of these 90 projects had partially achieved their output and result objectives. The Commission notes that 86 projects out of the 90 examined which have defined output and result indicators have reached (either fully or partially) or exceeded their corresponding targets. The Commission notes that, in the first place, it is the Monitoring Committees that are required to ensure that projects with EU added value are selected and approved. Performance evaluation of programmes will be reinforced for the 2014-2020 period: as foreseen in the 2014-2020 regulatory framework (Article 22(6) and (7) of Regulation (EU) No 1303/2013), the Commission will be able to sanction Member States in case of serious underachievement of a priority axis in a programme, as a result of the performance review (Article 22(6) of Regulation (EU) No 1303/2013) and at closure (Article 22(7)). Example of a project which fully achieved its objectives The project met its output and result objectives: An ESF project in Romania consisted of the organisation and delivery of training courses in the field of construction in order to provide participants with new skills that would improve their employment prospects. The grant agreement of the project defined five output and eight result indicators which were in line with the OP indicators at priority axis level. The project achieved all indicators within the deadline set in the grant agreement. The project even exceeded its targets in terms of participation of long-term unemployed women and youngsters. The remaining four projects (4 %) did not achieve any of their objectives. One of those four projects was withdrawn by the national authorities following our audit (see Figure 6.12). The Commission uses all possible measures at its disposal to ensure sound and efficient financial management in the use of EU funds. Moreover, it is only at the stage of the closure of the programme that the performance of projects financed by Operational Programmes will be evaluated. Example of a project which did not achieve its objectives For an ERDF project in Poland related to the development of a new method for burning brown coal, the beneficiary’s technical capability to undertake the project relied on the involvement of external contractors. Some of these contractors did not finally participate to the project. This contributed to the project not achieving its intended objectives. The intermediate body decided to cancel the project following our audit. Result indicators were not defined for 38 % of the projects For 57 of the 149 (38 %) completed projects the Member State authorities implemented a performance measurement system to allow the monitoring of project outputs but did not define any result indicators or targets in the project’s approval documents. For these projects, we could not measure if and to what extent they contributed to the achievement of the OP result objectives. The Commission and Member States assess to which extent OP result objectives have been achieved primarily through ex post evaluations, building on the monitoring and reporting of a series of output and result indicators. Our assessment in 15 Member States sampled showed that such cases were identified in 13 of them. However, the issue is particularly prevalent in Czech Republic, Spain, France, Italy, Latvia and Austria for ERDF/CF and Czech Republic, Germany, Spain and Poland for ESF. In these countries result indicators were not defined for more than half of the completed projects examined. However, 50 of these 57 completed ERDF/CF and ESF projects (88 %) had fully achieved their output objectives set by the managing authorities. The remaining seven (12 %) projects had partially achieved their output objectives. As regards the 2014-2020 programming period, we note that there is no legal requirement in the common provision regulation for Member States to define result indicators at project level. Nevertheless, we consider it good practice that managing authorities and intermediate bodies should whenever possible define relevant result indicators at project level that measure the contribution of the project to the achievement of the objectives set for the priority axis of the OP. The Commission notes that there is a major improvement for 2014-2020 with a comprehensive system of obligatory indicators for outputs and results at programme level. The achievement of the objectives is measured by a result indicator (with a baseline — the starting point — and target). Programmes will systematically measure and report progress of result indicators. Projects within a programme are selected to deliver outputs and contribute towards the results to be achieved at programme level. Their progress is measured with output indicators. These are also mandatory.

ANNEX

Anx. 61. 2015 2014 SIZE AND STRUCTURE OF THE SAMPLE Total transactions 223 331 ESTIMATED IMPACT OF QUANTIFIABLE ERRORS Estimated level of error 5,2 % 5,7 % Upper Error Limit (UEL) 7,6 % Lower Error Limit (LEL) 2,8 %

ANNEX 6.1 RESULTS OF TRANSACTION TESTING FOR ECONOMIC, SOCIAL AND TERRITORIAL COHESION

2015 2014 SIZE AND STRUCTURE OF THE SAMPLE Total transactions 223 331 ESTIMATED IMPACT OF QUANTIFIABLE ERRORS Estimated level of error 5,2 % 5,7 % Upper Error Limit (UEL) 7,6 % Lower Error Limit (LEL) 2,8 %

Anx. 62. E = DG Employment, Social Affairs and Inclusion; R = DG Regional and Urban Policy; X = Common assessment for both DGs Year Court Recommendation Court's analysis of the progress made Commission reply Fully implemented Being implemented Not implemented N/A under the current framework Insufficient evidence In most respects In some respects 2013 Chapter 5, recommendation 1: require from the Member States in their management declarations (according to Article 59(5)(a) of the Financial Regulation) an explicit confirmation regarding the effectiveness of the first level checks performed by the managing and certifying authorities. R Chapter 5, recommendation 2: carry out an assessment of the first level checks performed during the 2007-2013 programming period in accordance with Article 32(5) of the Financial Regulation. Taking account of the weaknesses identified, the Commission should analyse the costs and benefits of possible corrective measures and take (or propose) appropriate action (such as the simplification of the applicable provisions, improvements in the control systems and re-design of the programme or delivery system). R The Commission considers that its supervision of first-level checks through targeted audits of more risky areas where weaknesses were identified contributed to improvements for 2007-2013, and that the key elements of the 2014-2020 reformed period should lead to a more error-safe environment. Chapter 5, recommendation 3: analyse the underlying reasons for the high number of cases of non-compliance with EU state aid rules. R Chapter 5, recommendation 4: analyse the reasons for the persistent delays in disbursement of EU funds through FEIs and take corrective measures accordingly. R Chapter 5, recommendation 5: confirm in the annual activity report (AAR) of the Directorate-General for Regional and Urban Policy that the Commission’s calculation of the residual error rate is based on accurate, complete and reliable information on financial corrections. In order to do so, the Commission should request audit authorities to certify the accuracy of the data on financial corrections reported by certifying authorities for each OP whenever it deems such action necessary. R 2013 Recommendation 6 (recommendation 6 of chapter 5 and recommendation 6 of chapter 6): consistently disclose in its annual activity report (AAR) the reasons for not making reservations (or making reservations with a lower financial impact) in those cases where this is due to exceptions to applicable Commission guidance or approved audit strategies. X Recommendation 7 (recommendation 1 of chapter 6): should follow-up with the Member States the weaknesses identified in the DG EMPL’s risk based thematic audit of management verifications. This would require strengthening the checks related to compliance with public procurement rules and other relevant sources of errors (costs not linked to the project or with no added value). E Recommendation 8 (recommendation 2 of chapter 6): confirm in its AARs that it has carried out appropriate checks to ensure that the residual error rate is based on accurate, complete and reliable information on financial corrections. In order to do so, the Commission should request audit authorities to certify the accuracy of the data on financial corrections reported by certifying authorities for each OP, whenever it deems such an action necessary. E Recommendation 9 (recommendation 3 of chapter 6): should ensure that the application of Article 78 and 130 of the CPR for the 2014-2020 programming period will be applied in a manner that precludes the accumulation of pre-financing in addition to the initial pre-financing payment (see example Box 6.4). E Recommendation 10 (recommendation 4 of chapter 6): should ensure that the Member State authorities in charge of managing structural funds address the issue of charging personnel costs at higher rates for EU projects compared to those financed by national funds. E Recommendation 11 (recommendation 5 of chapter 6): ensures, when approving the OPs for the new programming period, that Member States have considered all simplification possibilities allowed by the 2014-2020 European Structural and Investment Funds regulations. E 2012 The recommendations issued in the context of the 2012 annual report that are substantially similar to those issued in 2013 have been analysed in conjunction. Recommendation 1 addresses weaknesses in first level checks carried out by managing authorities and intermediate bodies for the ERDF and the CF, through specific guidance material and, where appropriate, through training measures. X Recommendation 2 on the basis of its experience gained of the 2007-2013 programming period, carries out an assessment of the use of national eligibility rules in view of identifying possible areas for further simplification and eliminating unnecessary complex rules (gold-plating). E R Chapter 5, recommendation 3: specifies clear rules and provides robust guidance on how to assess the eligibility of projects and calculate the co-financing for revenue- generating ERDF and CF projects under the 2014-2020 programming period. R Recommendation 4 (recommendation 3 of chapter 6): promotes the extensive use of simplified cost options with a view to reducing the risk of error in cost declarations and the administrative burden on beneficiaries. The flat rates for simplified cost options should be systematically approved/validated in advance by the Commission to ensure that they meet the regulatory requirements (fair, equitable and verifiable calculation). E 2012 Recommendation 5: seeks improvement in the work done by audit authorities and the quality and reliability of the information provided in ACRs and audit opinions. X The Commission notes the general improvement recognised by the Court in the checks carried out by the audit authorities, in particular on public procurement and State aid rules (see paragraph 6.59). The Commission also refers to its reply to paragraphs 6.45 and 6.52.

ANNEX 6.2 FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR ECONOMIC, SOCIAL AND TERRITORIAL COHESION

E = DG Employment, Social Affairs and Inclusion; R = DG Regional and Urban Policy; X = Common assessment for both DGs Year Court Recommendation Court's analysis of the progress made Commission reply Fully implemented Being implemented Not implemented N/A under the current framework Insufficient evidence In most respects In some respects 2013 Chapter 5, recommendation 1: require from the Member States in their management declarations (according to Article 59(5)(a) of the Financial Regulation) an explicit confirmation regarding the effectiveness of the first level checks performed by the managing and certifying authorities. R Chapter 5, recommendation 2: carry out an assessment of the first level checks performed during the 2007-2013 programming period in accordance with Article 32(5) of the Financial Regulation. Taking account of the weaknesses identified, the Commission should analyse the costs and benefits of possible corrective measures and take (or propose) appropriate action (such as the simplification of the applicable provisions, improvements in the control systems and re-design of the programme or delivery system). R The Commission considers that its supervision of first-level checks through targeted audits of more risky areas where weaknesses were identified contributed to improvements for 2007-2013, and that the key elements of the 2014-2020 reformed period should lead to a more error-safe environment. Chapter 5, recommendation 3: analyse the underlying reasons for the high number of cases of non-compliance with EU state aid rules. R Chapter 5, recommendation 4: analyse the reasons for the persistent delays in disbursement of EU funds through FEIs and take corrective measures accordingly. R Chapter 5, recommendation 5: confirm in the annual activity report (AAR) of the Directorate-General for Regional and Urban Policy that the Commission’s calculation of the residual error rate is based on accurate, complete and reliable information on financial corrections. In order to do so, the Commission should request audit authorities to certify the accuracy of the data on financial corrections reported by certifying authorities for each OP whenever it deems such action necessary. R 2013 Recommendation 6 (recommendation 6 of chapter 5 and recommendation 6 of chapter 6): consistently disclose in its annual activity report (AAR) the reasons for not making reservations (or making reservations with a lower financial impact) in those cases where this is due to exceptions to applicable Commission guidance or approved audit strategies. X Recommendation 7 (recommendation 1 of chapter 6): should follow-up with the Member States the weaknesses identified in the DG EMPL’s risk based thematic audit of management verifications. This would require strengthening the checks related to compliance with public procurement rules and other relevant sources of errors (costs not linked to the project or with no added value). E Recommendation 8 (recommendation 2 of chapter 6): confirm in its AARs that it has carried out appropriate checks to ensure that the residual error rate is based on accurate, complete and reliable information on financial corrections. In order to do so, the Commission should request audit authorities to certify the accuracy of the data on financial corrections reported by certifying authorities for each OP, whenever it deems such an action necessary. E Recommendation 9 (recommendation 3 of chapter 6): should ensure that the application of Article 78 and 130 of the CPR for the 2014-2020 programming period will be applied in a manner that precludes the accumulation of pre-financing in addition to the initial pre-financing payment (see example Box 6.4). E Recommendation 10 (recommendation 4 of chapter 6): should ensure that the Member State authorities in charge of managing structural funds address the issue of charging personnel costs at higher rates for EU projects compared to those financed by national funds. E Recommendation 11 (recommendation 5 of chapter 6): ensures, when approving the OPs for the new programming period, that Member States have considered all simplification possibilities allowed by the 2014-2020 European Structural and Investment Funds regulations. E 2012 The recommendations issued in the context of the 2012 annual report that are substantially similar to those issued in 2013 have been analysed in conjunction. Recommendation 1 addresses weaknesses in first level checks carried out by managing authorities and intermediate bodies for the ERDF and the CF, through specific guidance material and, where appropriate, through training measures. X Recommendation 2 on the basis of its experience gained of the 2007-2013 programming period, carries out an assessment of the use of national eligibility rules in view of identifying possible areas for further simplification and eliminating unnecessary complex rules (gold-plating). E R Chapter 5, recommendation 3: specifies clear rules and provides robust guidance on how to assess the eligibility of projects and calculate the co-financing for revenue- generating ERDF and CF projects under the 2014-2020 programming period. R Recommendation 4 (recommendation 3 of chapter 6): promotes the extensive use of simplified cost options with a view to reducing the risk of error in cost declarations and the administrative burden on beneficiaries. The flat rates for simplified cost options should be systematically approved/validated in advance by the Commission to ensure that they meet the regulatory requirements (fair, equitable and verifiable calculation). E 2012 Recommendation 5: seeks improvement in the work done by audit authorities and the quality and reliability of the information provided in ACRs and audit opinions. X The Commission notes the general improvement recognised by the Court in the checks carried out by the audit authorities, in particular on public procurement and State aid rules (see paragraph 6.59). The Commission also refers to its reply to paragraphs 6.45 and 6.52.

CHAPTER 7 Natural resources

Txt. 1. THE COURT’S OBSERVATIONS INTRODUCTION This chapter presents our findings for Natural resources. The chapter presents separately the results of testing of the European Agricultural Guarantee Fund (EAGF) and for other spending (Rural development, environment, climate action and fisheries) under the multi-annual financial framework heading (MFF). Key information on the composition of this MFF heading is provided in Figure 7.1. MFF heading 2 Natural resources — Key information 2015 (billion euro) Source: 2015 consolidated accounts of the European Union. Total payments for the year 58,0 - advances 2,4 + clearings of advances 2,9 Audited population, total 58,5 THE COURT’S OBSERVATIONS In addition to our findings with regard to regularity for the specific assessments, this chapter presents, in a separate section, the results of some performance-related issues for a sample of rural development projects. Brief description of Natural resources The common agricultural policy (CAP) is the basis for EU spending on agriculture. The Commission, in particular the Directorate-General for Agriculture and Rural Development (DG AGRI), and paying agencies in the Member States share the management of implementation of the CAP (see paragraph 7.31). They pay beneficiaries through two funds: the EAGF, which fully finances EU direct aid and market measures, and the European Agricultural Fund for Rural Development (EAFRD), which co-finances rural development programmes together with the Member States. The main measures financed by the EAGF are: (1) The direct aid Single Payment Scheme (SPS), 29,3 billion euro in 2015, and Single Area Payment Scheme (SAPS), 7,8 billion euro in 2015. These area-related schemes provide decoupled income support. (2) Other direct aid schemes, 5,1 billion euro in 2015, and interventions in agricultural markets, 2,7 billion euro in 2015. Txt. 2. The EAFRD co-finances the rural development expenditure disbursed through Member States' rural development programmes. The expenditure covers both area-related measures and non-area-related measures. In 2015, the EAFRD accounted for 11,8 billion euro of expenditure. The main risks to regularity are that: (3) direct aid is paid for ineligible land or to ineligible beneficiaries; (4) aid for interventions in agricultural markets is granted for ineligible costs or to ineligible applicants; (5) rural development expenditure is ineligible, due to non-compliance with often complex rules (including procurement) and eligibility conditions, in particular for investment measures. Txt. 3. This chapter also covers the EU environmental policy, climate action, and the common fisheries policy. The main risk to regularity in these areas is that aid is granted for ineligible or overstated costs. Audit scope and approach Annex 1.1 sets out the key elements of our audit approach and methodology. For the audit of Natural resources, the following specific points should be noted: (a) We examined a sample of 180 transactions for EAGF and 179 transactions for Rural development, environment, climate action and fisheries as defined in Annex 1.1, paragraph 7. Each sample is designed to be representative of the range of spending under the MFF heading for each of the two specific assessments. In 2015 the EAGF sample consisted of transactions from 18 Member States. For the second specific assessment the sample consisted of 159 transactions for rural development and 20 other transactions including environment, climate action and fisheries from 17 Member States. (b) We assessed selected internal control systems through a desk review of 24 Commission audits covering the EAGF and the EAFRD. We complemented this work by re-performing six of the Commission’s audits in the Member States concerned. We performed additional audit work in these six Member States for a limited number of risk-prone areas relating to the EAGF and the EAFRD, as well as for cross-compliance. We also reviewed five Commission audits covering the common fisheries policy, and re-performed one. Txt. 4. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES (c) We assessed the annual activity reports of DG AGRI and DG MARE. Txt. 5. For both the EAGF and the EAFRD, 2015 was the last year when most payments were made to final beneficiaries on the basis of the rules prevailing before the 2013 CAP reform. PART 1 — OUR ASSESSMENT ON REGULARITY Regularity of transactions Annex 7.1 contains a summary of the results of transaction testing for Natural resources as a whole as well as for each of the two specific assessments (EAGF and Rural development, environment, climate action and fisheries). Of the 359 transactions examined, 121 (34 %) were affected by error. On the basis of the 87 errors we have quantified, the estimated level of error for Natural resources is 2,9 %. The Commission takes note of the most likely error estimated by the Court which is broadly the same as last year excluding cross-compliance. The Commission is of the view that net financial corrections resulting from multiannual conformity procedures, as well as recoveries from beneficiaries reimbursed to the EU budget, represent a corrective capacity that has to be taken into account in a comprehensive assessment of the overall system of internal control. The Commission also notes that, as reported in the 2015 annual activity report of the Directorate General for Agriculture and Rural Development, its corrective capacity amounted in 2015 to 1065,710 million euro (1,87 % of the total CAP expenditure). Figure 7.2 presents the breakdown of the different types of errors to our overall estimated level of error for 2015. Breakdown of the estimated level of error — Natural resources Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The nature and pattern of errors is significantly different between the two specific assessments, as set out in the following paragraphs. The Commission notes that there are significant differences in the design, scope and objectives of both pillars of the CAP which explain these differences in the pattern of CAP-related errors. EAGF — Market and direct support Annex 7.1 contains a summary of the results of transaction testing. For the EAGF, of the 180 transactions that we examined, 52 (29 %) were affected by error. On the basis of the 47 errors we have quantified, the estimated level of error is 2,2 %. The Commission takes note of the most likely error estimated by the Court which is broadly the same as last year excluding cross-compliance. The Commission also notes that the frequency of errors has steadily decreased over the last three years. Figure 7.3 presents the breakdown of the different types of errors to our estimated level of error for 2015. Breakdown of the estimated level of error — EAGF Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES In four cases of quantifiable errors made by final beneficiaries, national authorities had sufficient information to prevent, or detect and correct the errors before declaring the expenditure to the Commission. If all this information had been used to prevent, or detect and correct errors, the estimated level of error for this specific assessment would have been 0,3 percentage points lower. In addition, we found that for another four cases, the error that we detected had been made by the national authorities. These errors contributed 0,6 percentage points to the estimated level of error. We give an example of errors made by the national authorities in Figure 7.5. The Commission welcomes the Court’s assessment on Member States’ role and is of the view that Member States should have done more to reduce errors. The Commission continues to work with Member States to reduce errors. Most of the quantifiable errors related to overstatements of the eligible area of land (see Figure 7.3). In 2015, we found such errors in 12 out of the 18 Member States visited. Around half of these quantifiable errors were smaller than 2 % (see paragraph 7.18) and thus had only a limited impact on the overall estimated level of error. A further quarter of these errors were between 2 % and 5 %. The Commission takes note that most of the over-declaration errors detected by the Court were relatively small errors and welcomes the Court’s assessment on the improvement of the reliability of LPIS data. The Commission considers that minor errors are impossible to avoid at a reasonable cost. The Commission encourages Member States to update their land parcel identification system (LPIS) on a regular basis to ensure detecting large over-declarations. The Commission considers the fact that the large majority of the errors reported by the Court were under 5 % to be proof of the good functioning of the integrated administrative control system (IACS) of which LPIS is a fundamental component. As land changes over time, differences between LPIS and the reality on the ground may occur, however regular updates of the system limit the risk. To promote frequent updates of the LPIS, the Commission has provided for, as a pre-condition to reduce on-the-spot checks from 5 % to 1 % for some direct payments schemes, an LPIS update at least every three years with ortho-images aged less than 15 months (the update covering per year should be at least 25 % of the eligible hectares). A quality assessment of LPIS was also introduced in 2010 and maintained in the present period (Regulation (EU) No 640/2014). This yearly exercise done by Member States’ administrations and monitored by the Commission is a quality control process whose ultimate objective is to improve the quality and reliability of the LPIS, by analysing detected technical weaknesses and by taking remedial actions to address them. The annual on-the-spot checks are still a way, on a sample basis, to achieve an accurate and updated LPIS. In the Member States, the key control tool for area aid is the Land Parcel Identification System (LPIS), which is a database containing eligibility data for all agricultural parcels. It contains computerised graphical information, including ortho-images (aerial photographs) which Member States should update at least every five years. Provided the graphical information is properly updated and analysed, the LPIS makes a significant contribution to preventing and detecting area-related errors in aid applications. Thanks to action plans set up by the Commission and the Member States, the reliability of data in the LPIS has been constantly improving over recent years. Nevertheless, we still detect area-related errors. This is because administrative checks on the eligibility of land are based on the same LPIS data which farmers frequently copy into their aid applications. Therefore, if the LPIS data is incorrect, such checks cannot normally detect eligibility errors in the claim. The technical standard required for ortho-images does not usually allow the farmers and the paying agencies to detect small errors (smaller than 2 %, see paragraph 7.16), which are, to a large extent, impossible to avoid at a reasonable cost. However, the larger area-related errors were, in general, detectable. In such cases, the eligible area recorded in the LPIS database and declared by the farmer was substantially incorrect, mainly because: (9) the paying agencies did not reduce payments sufficiently, or at all, for ineligible vegetation and features visible on the ortho-images of the land parcel concerned when assessing the eligible area for the purpose of updating the LPIS, or (10) national definitions of eligible permanent pasture did not comply with the EU definition. Txt. 6. Whenever there is a case of national definition for permanent pasture that did not comply with the EU definition, specific corrective actions including, if necessary, action plans are set up to address the issue. To protect the EU budget, the Commission also applies reductions and suspensions of payments and net financial corrections in the framework of conformity clearance procedures whenever necessary. We present an example of incorrect eligibility data in the LPIS in Figure 7.4. Example of irregular payments due to incorrect eligibility data in the LPIS Permanent pasture is defined as land used to grow grasses or other herbaceous forage. In our 2014 annual report we reported that in Spain, despite considerable remedial action, important weaknesses in the LPIS relating to the eligibility of permanent pasture still persisted, and that in 2014 and 2015 further corrective action was initiated. We inspected a parcel of 48 hectares in Spain (Madrid), recorded in the LPIS as fully eligible permanent pasture at the time the farmer made the claim. The ortho-image showed however the presence of dense bushes and trees. In his 2014 aid application, the farmer had declared the parcel on the basis of the eligible area recorded in the LPIS and he was paid SPS aid for the entire parcel. We found the parcel to be ineligible for SPS aid. In 2015 the Spanish authorities carried out a review of their LPIS on the basis of new ortho-images and using new technology, and recorded the parcel concerned as ineligible. We also found that area aid was claimed and paid on the basis of incorrect eligibility information in the LPIS, which was detectable on the ortho-images, in Greece, Spain (Aragon, Castilla-La Mancha, Castilla y León and Extremadura), Italy (Calabria) and the UK (England). Example of irregular payments due to incorrect eligibility data in the LPIS Issues around permanent pasture have been identified by the Commission before 2014 and action plans have been implemented in the Member States concerned. In Spain, the national authorities carried out full re-measurement of all the parcels declared by the beneficiaries audited by the Court and launched recovery procedures where necessary. Furthermore, the fact that the parcel under consideration was recorded as ineligible in 2015 proves that the procedure to update LPIS in Spain works. Audits carried out by the Commission services in previous years had identified deficiencies linked to the eligibility of permanent pasture in Greece and Spain and ensured that they were addressed through specific action plans which led to improvements (see Court’s 2014 annual report, box 7.8). In both Greece and Spain, the area recorded as eligible in LPIS was significantly reduced: in Greece the eligible permanent pasture area was reduced from 3,6 million hectares in 2012 to 1,5 million hectares in claim year 2014 and in Spain, the eligible permanent pasture area was reduced from 18,5 million hectares in 2013 to 6,4 million hectares in 2016. For the UK (England) the remedial action plan is ongoing and is being closely monitored by the Commission. Furthermore, conformity clearance procedures ensure that the risk to the EU budget is adequately covered by net financial corrections. We present an example of national eligibility criteria not complying with EU legislation in Figure 7.5. Example of national eligibility criteria not complying with EU legislation In our 2013 annual report we reported that French authorities grant aid for grazable heathland, although these areas comprise a mixture of eligible herbaceous vegetation and ineligible bushes, dense woods, etc. In 2015 we found six such parcels to be fully or partially ineligible because they contained dense trees and bushes, which the French authorities had assessed as fully eligible. Example of national eligibility criteria not complying with EU legislation Audits carried out by the Commission prior to 2013 had already identified similar deficiencies in France. The weaknesses are being followed up under a specific action plan established by the Member State and closely monitored by the Commission. Any such weaknesses that are found are followed up through conformity clearance procedures which ensure that the risk to the EU budget is adequately covered by net financial corrections. In parallel, the Commission remains vigilant that the action plan is adhered to in order to ensure improvement. In light of certain delays incurred further measures pursuant to article 41(2) of Regulation (EU) No 1306/2013 were initiated as of 2015. Rural development, environment, climate action and fisheries Annex 7.1 contains a summary of the results of transaction testing. For Rural development, environment, climate action and fisheries, of the 179 transactions that we examined, 69 (39 %) were affected by error. On the basis of the 40 errors we have quantified, the estimated level of error is 5,3 %. The Commission takes note of the most likely error estimated by the Court, which is lower than that observed last year. Figure 7.6 presents the breakdown of the different types of errors to our estimated level of error for 2015. Breakdown of the estimated level of error — Rural development, environment, climate action and fisheries Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES For rural development, of the 159 transactions we examined, 78 were area-related and 81 non-area-related. Of these 159 transactions, 64 (40 %) were affected by error, of which 38 were quantifiable. As regards environment, climate action and fisheries, of 19 transactions sampled, 5 (26 %) were affected by error, of which 2 were quantifiable. For the one remaining transaction, we found no error. For rural development, the frequency of errors (both quantified and not-quantified) found by the Court has steadily decreased over time. In 13 cases of quantifiable errors made by final beneficiaries, national authorities had sufficient information to prevent, or detect and correct the errors before declaring the expenditure to the Commission. If all this information had been used to prevent, or detect and correct errors, the estimated level of error for this specific assessment would have been 1,7 percentage points lower. In addition, we found that for six cases, the error that we detected had been made by the national authorities. These errors contributed 1,5 percentage points to the estimated level of error. We give an example of an error made by the national authorities in Figure 7.7. The Commission is aware that the national authorities could potentially have detected many of the errors found by the Court. The CAP rules provide the Member States with all necessary instruments to mitigate most of the risk of errors. For the example given in Figure 7.7, the Commission has launched in March 2016 a conformity clearance procedure. All necessary actions will be taken to protect the EU budget. The types of errors are similar to those we found in previous years and to those identified in our special report analysing the causes of errors in rural development spending. As in previous years, non-area related measures were the major component (82 %) of the estimated level of error reported in paragraph 7.21. The Commission is aware of the root causes of errors and is working with the Member States to remedy the situations by offering guidance or requesting action plans to be implemented. For instance, the Commission knows that non-area-related measures are more prone to errors. For that reason systemic solutions are promoted, such as simplifying the design of measures or using less complex calculation methods, for example simplified cost options or lump sums. The Commission is of the view that irregularities in the public procurement procedure do not necessarily mean that the overall expenditure was ineligible. In order to assess the financial risk of public procurement errors, the Commission adopted in December 2013 guidelines for determining financial corrections for non-compliance with public procurement rules in projects implemented under the shared management Of the 81 non-area related transactions examined, 15 (19 %) did not comply with the eligibility requirements (an example is given in Figure 7.7). For environment, climate action and fisheries, the 2 quantifiable errors were due to ineligible expenditure. Example of ineligible expenditure or activity The rural development measure animal welfare payments provides support to farmers who voluntarily make animal welfare commitments. Payments are made annually and cover additional costs and income foregone resulting from the commitment made, such as increased space for animals. This is an example of using simplified cost options, which may reduce the administrative burden for Member States and beneficiaries. Such options should be based on a sound methodology, in particular an accurate calculation of the simplified cost. In Romania we found ineligible expenditure because the national authorities, when calculating the simplified costs, did not appropriately take into account the number of production cycles. As a result, payments to all beneficiaries were systematically overstated. We checked all types of commitments for the animal welfare payments in Romania, including those outside the sample, in order to determine to what extent the miscalculations affected the entire expenditure for animal welfare in Romania. We found that total ineligible expenditure from the time the measure started (2012) to October 2015 could be up to 152 million euro (out of 450 million euro in total). We also found eligibility errors in Belgium (Wallonia), Germany (Rhineland-Palatinate), Spain (Andalucia, Galicia, Castilla-La Mancha), Greece, Italy (Rural network, Sicily and Campania) and Poland. Example of ineligible expenditure or activity The Commission fully supports Member States in their efforts to use whenever possible simplified costs options as a tool to prevent errors, as recognised by the Court in its special report No 23/2014 (see paragraph 86). As regards the Romanian cases, the Commission agrees that there was an error in the calculation of the simplified cost used for the payments. Following the Court’s findings, the Commission launched a conformity clearance procedure in 2016 to establish precisely the amount at risk. This may result in the application of a financial correction. For 27 transactions examined, the beneficiaries had to comply with public procurement rules. These rules ensure purchase of goods and services on the most favourable terms, while guaranteeing equal access to public contracts and compliance with the principles of transparency and non-discrimination. In six cases one or more of these rules had been breached. We present an example of such an error in Figure 7.8. The Commission has identified public procurement as an important source of errors and intensified its audit work to ensure appropriate protection of the EU budget. Compliance with public procurement rules is also one of the central elements of the action plans for rural development. In-depth audits of that area were carried out in 2015/2016 in four Member States. The practice of in-depth checks on public procurement rules continues in other Member States in 2016. In order to assess the financial risk of public procurement errors, the Commission adopted in December 2013 guidelines for determining financial corrections for non-compliance with public procurement rules in projects implemented under the shared management. Example of non-compliance with public procurement rules A commune in Italy (Campania) received support for the construction of a monorail: infrastructure to help farmers transport their olives up and down the hills. The contract was awarded to a private company through a public procurement procedure. In order to assess the various offers, the commune established a set of criteria, with different weightings (measured in points). The contract notice required the bidders to include, as part of their offers, proposals for additional works concerning new roads, sewage and an aqueduct. These additional works accounted for 75 % of the total points on which the commune based the award decision. By comparison, price only accounted for 10 % of the total points. Since the EU aid was granted for the construction of the monorail, the beneficiary breached Italian public procurement law by setting up criteria — the construction of roads, sewage and an aqueduct — which were not related to the subject, nature and characteristics of the contract notice. Such criteria were therefore irregular under Italian public procurement rules. We also found cases of non-compliance with public procurement rules in Germany (Rhineland-Palatinate), Greece and Romania. Example of non-compliance with public procurement rules The Commission understands that the award criteria as set in the call for tender were not fully related to the nature of the works. However, the Commission underlines that the investment was carried out as planned. The Italian authorities reviewed the file and instigated recovery procedure for breaches in the public procurement procedures. The Italian authorities applied a 10 % financial correction. The Commission considers that this corrective action helps protect the EU budget. Reservations were made in the 2015 annual activity report to cover public procurement issues, as recognised by the Court in point 7.60. Our sample included 45 transactions for agri-environment payments, which concerned the use of agricultural production methods compatible with the protection of the environment, landscape and natural resources. We found that in three cases farmers had not met all the conditions for payment. The Commission would like to emphasise that all three beneficiaries had a wide range of different agri-environmental commitments to observe and implemented successfully the majority of them. Only few of the commitments were not respected fully or not properly documented. Of the 78 area-related transactions, we found 15 cases (19 %) where the eligible area of land was overstated. Most of the overstatements we identified were relatively small, and thus had only a limited impact on the overall estimated level of error. See also paragraph 7.18. The Commission takes note that most of the over-declaration errors detected by the Court were relatively small errors and welcomes the Court’s assessment on the improvement of the reliability of LPIS data (see paragraph 7.17). The Commission considers that minor errors are impossible to avoid at a reasonable cost. National authorities should check the reasonableness of costs. In 24 investment projects, the national authorities had not checked this adequately. We reported these cases as errors. However, it is generally not possible to determine the amount of ineligible expenditure, in which case we cannot quantify the error. We include additional information on the reasonableness of costs in the section testing of performance criteria (see paragraph 7.71). See Commission’s reply to point 7.71. Examination of annual activity reports and other elements of internal control systems DG AGRI’s systems related to the regularity of transactions in agriculture The Commission and Member States share the management of almost all agricultural expenditure. Some 80 paying agencies in the Member States disburse the aid to farmers and other beneficiaries; the Commission reimburses the Member States for these costs each month (for the EAGF) or each quarter (for the EAFRD). The Commission carries out conformity audits aimed at assessing whether Member States’ systems ensure that aid payments to beneficiaries are legal and regular. Systems weaknesses detected may lead to financial corrections, imposed on Member States by way of conformity decisions. Financial corrections are often made on a flat-rate basis and cover expenditure relating to several budgetary years. They rarely result in recoveries of payments from final beneficiaries and are thus paid from national budgets. While it is true that the majority of corrections in the past were flat rate, in the new legal framework the Member States are encouraged to provide a calculation for the risk to the Fund. Where undue payments, which can be directly linked to a beneficiary, are identified as a result of the Commission’s conformity clearance procedures, Member States are required to launch the necessary recovery actions against the final beneficiaries. To check that the Commission’s conformity audits complied with international auditing standards and the main regulatory provisions, we carried out desk reviews of a sample of 24 out of 120 conformity audits reported by DG AGRI in 2015. We also re-performed six of these 24 audits. In 2015, the Commission updated its audit procedures following a self-assessment of its audit methodology and quality assurance policy. We reviewed the updated procedures and took them into account for our assessment. The Commission’s conformity audits are overall compliant with the relevant standards. However, when the Commission does not raise any findings, it needs to improve the way it documents the audit work supporting such conclusion. The Commission strives to respect ISSAI when performing its audit. This is one of the reasons that led it to use the services of an external company to support its self-assessment of its audit procedures. The audit reports of each mission are the basis for the findings raised by the Commission. When no findings are raised (e.g. there was no issue to report, the weaknesses detected were not material), this is stated in the audit report. In previous annual reports, we have drawn attention to the problem that conformity procedures were too lengthy, resulting in a significant backlog of open files. In 2015, the Commission significantly reduced the number such old open files: from 192 to 34. Changes in legislation for the 2014-2020 period, which took effect for the CAP on 1 January 2015, were designed to streamline the conformity clearance procedure by introducing legal deadlines. As a result, the Commission now monitors the audit cycle more closely in order to comply with the internal and external deadlines. The Commission follows up closely the closure of backlog audits (old open files): their number has been further reduced from 34 to 20 in June 2016. The Commission’s objective is to close all remaining backlog audits (old open files) by the end of 2016. The new legislation for the 2014-2020 period also placed greater emphasis on the Member State’s role in establishing the amount at risk, which the Commission may use to calculate the financial correction to be imposed. We reviewed whether the Commission correctly validated and used the Member States’ data by examining the Commission guidance and checking six financial corrections adopted or prepared in 2015. As recommended by the Court in previous years, the Commission has developed a framework that is intended to increase the number of calculated financial corrections and thereby apply less flat rate corrections. Taking into account the inherent difficulty in cases where numerous transactions are concerned, appropriate legal provisions have been inserted in the basic act, including the principles of Member States’ responsibility to provide the necessary information and of proportionate effort. The delegated act contains detailed provisions on the calculated and extrapolated financial corrections. The Commission’s detailed guidelines on financial corrections have been updated accordingly, notably regarding the criteria and methodology to be used by the Commission to accept or not the Member States’ calculation or extrapolation of the risk to the EU budget. The Commission considers that the audit manual includes information on the verification of data supplied by Member States. Communication C(2015) 3675 attached to the audit manual sets out clearly the minimum requirements that Member States have to meet when they submit additional data to the Commission for the calculation of the financial corrections and the criteria and methodology to be used by the Commission to accept or not the calculation or extrapolation of the financial correction. It is standard procedure for the Commission to do thorough checks on the data provided by Member States. This is indeed confirmed by the findings of the Court for four of the six financial corrections examined. As regards the two other financial corrections mentioned: (11) for one, the amount involved was relatively small and additional information was requested from the Member State, (12) for the 5,8 million euro financial correction, the Commission checked the validity of the data given by the Member State although this process was not properly documented. Txt. 7. The Commission would like to underline that the efforts to check the figures provided by Member States should be proportionate to the potential financial consequences. The Commission can also rely on the work of the certification bodies. The Commission audit manual does not include detailed audit procedures and documentation requirements for the verification of the data supplied by the Member States. For four of the six financial corrections examined, we found that the Commission carried out audit visits to verify the data received from the Member States concerned. For the other two financial corrections the Commission did not provide us with sufficient and appropriate evidence of their checks or verifications. For one of these cases, the financial correction was relatively small. However, for the remaining case, the financial correction was 5,8 million euro. The use of non-verified data creates a financial risk to the EU budget, as it is in the Member States’ interest to reduce financial corrections. Member States’ systems related to the regularity of transactions in agriculture Member States' authorities are responsible for putting in place and operating: Pt_a_. 2. appropriate administrative and control procedures to ensure the accuracy of declarations made by claimants and the fulfilment of eligibility requirements; Pt_b_. 2. on-the-spot checks which, depending on the aid scheme, should cover at least 5 % of all beneficiaries or of the relevant expenditure. Txt. 8. We based our systems examination on audits carried out by the Commission (DG AGRI). Within the sample of 24 conformity audits we reviewed (see paragraph 7.32), the Commission reported significant weaknesses for 17 of the systems concerned. Since we found that the Commission’s work was overall satisfactory (see paragraph 7.33), we can rely on the Commission’s assessment of these systems. The Commission welcomes the Court’s statement that the Commission’s work was overall satisfactory and that it can rely on the Commission’s assessment of Member States’ systems. In the six Member States we visited (see paragraph 7.32), we also performed additional audit work for a limited number of risk-prone areas. Our findings for the EAGF, rural development, and cross-compliance are set out in the following paragraphs. Member States' systems related to regularity of EAGF transactions The Integrated Administration and Control System (IACS) is the main management and control system for direct aid payments. The system consists of databases of farm holdings and aid applications, a Land Parcel Identification System (LPIS; see also paragraphs 7.17 and 7.18), animal databases and a database of entitlements in the Member States implementing the SPS. The paying agencies carry out administrative cross-checks between these databases with the objective of ensuring that payments are made for the correct amount, to eligible beneficiaries for eligible land or animals. The Commission considers that IACS, including LPIS, is a reliable system and that the quality and coverage of IACS is essential for assuring the regularity of direct aid payments and EAFRD expenditure, in particular agri-environmental measures and less-favoured areas. Whenever weaknesses occur, the Member States must put in place corrective actions, including specific action plans, to address them. These are carefully monitored by the Commission. In our 2014 annual report, we stated that IACS makes a significant contribution in preventing and reducing the levels of error in the aid scheme to which it applies. This year, we examined the reliability of selected key control functions in the IACS at the six paying agencies visited (see paragraph 7.32). Without calling into question IACS contribution in preventing and reducing the levels of error, we found that some of the key control functions we examined were affected by weaknesses. These related to the LPIS, the administrative controls, the quality of on-the-spot inspections and the recovery procedures for incorrect payments. We provide an example of the weaknesses we identified in Figure 7.9. Example of the weaknesses found concerning the LPIS For six consecutive years (2010-2015) in the United Kingdom (Northern Ireland) the annual LPIS quality assessment showed that it failed to comply with several quality elements required by EU legislation. Despite the legal obligation to put in place remedial action in such situations, the quality assessment for 2015 showed that there was no consistent improvement: overall, the eligible areas recorded in the Northern Ireland LPIS were overstated by more than 2 %, leading to a risk of irregular payments. We have already reported the weaknesses in the Northern Ireland LPIS in our 2012 annual report (Annex 3.2). Example of the weaknesses found concerning the LPIS From 2013, corrective actions were taken by the Northern Irish authorities following audits carried out by the Commission. The estimation of eligible area in the LPIS quality assessment (QA) was non-conforming for 2010, 2011, 2012 and 2015. The reasons were explained by the UK-NI authorities in the respective annual QA reports. In the years when the 3-year LPIS refresh started (2013, 2014), the number of non-conforming parcels decreased. However, in 2015, the UK-NI authorities detected that there was an increase of the number of parcels with no eligible area which had not been identified through the refresh programme. The Commission continues to closely monitor the situation. Member States' systems related to regularity of transactions in rural development For five out of the six paying agencies visited (see paragraph 7.32), we identified deficiencies in administrative checks related to eligibility conditions, in particular those concerning public procurement. These system weaknesses were very similar to those found and reported in previous years. System failures are an important cause of the errors detected during transaction testing (see paragraph 7.24). The Commission carries out conformity audits in the Member States to verify that the expenditure paid is in compliance with the rules. In these audits, the Commission also found weaknesses in administrative checks related to eligibility conditions and public procurement. Corrective actions have been implemented, including when necessary action plans. In addition reservations were made in the 2015 annual activity report. Financial corrections have also been made in this respect and a number of conformity clearance procedures are currently ongoing and are likely to lead to further financial corrections. A dedicated audit of public procurement was launched in 2015 in 4 Member States. Other Member States will be audited in 2016. For public procurement, we examined a sample of 41 projects (in addition to our transaction testing), of which eight (20 %) were affected by errors. In three out of the eight cases, we identified ineligible expenditure, amounting in total to more than 1,5 million euro of EAFRD aid. In an effort to reduce the level of error in rural development spending, the Commission and the Member States are implementing action plans to identify the causes of error and to implement targeted corrective action to reduce the risk of errors occurring. We reviewed whether, for frequent errors, the action plan of the Member State concerned dealt with the root cause of these errors. We looked at 14 of the most significant quantifiable errors detected in eight Member States during our 2015 transaction testing. We noted that seven out of the eight had already updated their action plans by February 2016, including measures to address some of the issues we found. During our examination of Member States’ systems, however, we found that the four paying agencies where we had detected weaknesses in the administrative checks on public procurement procedures had not included any action on this issue in their action plans. While we consider that the action plans are generally a sound tool to address the causes of errors, there is still scope for further improvement in the area of public procurement. When weaknesses are identified in the framework of conformity audits and audits by the Court, the Commission always requests that Member States take corrective actions, which may take the form of an action plan. Following the Court’s recommendations in special report No 23/2014, the Commission has recommended all Member States to pay particular attention to the issue of public procurement and, where necessary, take remedial actions. Reservations were made in the 2015 annual activity report to cover public procurement issues, as recognised by the Court in point 7.60. Member States' systems related to cross-compliance We examined the system for ensuring that cross-compliance requirements are met in the six paying agencies visited (see paragraph 7.32). In particular, we analysed the reliability of control statistics and the procedures put in place to ensure that Member States appropriately check cross-compliance requirements for protecting waters against nitrates (the Nitrates Directive) and for cattle notifications. We also reviewed 63 control reports. The Commission notes that cross-compliance does not relate to assessment of the legality and regularity of underlying transactions. We found deficiencies, mainly relating to the reliability of control statistics and sampling, in five out of the six Member States visited. We found problems (e.g. incorrect calculation of sanctions or lack of checks on certain requirements) in 13 of the 63 control reports reviewed. Moreover, in Lithuania and the UK (Northern Ireland) in four of the 24 control reports sampled the sanctions were not correctly included in the control statistics. We give another example of the deficiencies we identified in Figure 7.10. The Commission shares the Court’s observation and pays particular attention to these requirements during its audits. In many of them, the Commission has observed weaknesses in the scope and quality of cross-compliance controls. When a systemic non-compliance is established, the Commission always requests corrective actions, including actions plans whenever necessary and ensures the protection of the EU budget via the conformity clearance procedure. Example of insufficient cross-compliance checks In Lithuania we found that national authorities did not sufficiently check the cross-compliance requirements for protecting waters against nitrates, which are applicable to all farmers. The Member State used several thresholds, under which farmers were not subject to cross-compliance checks for requirements resulting from the Nitrates Directive. For example, farmers with a utilised agricultural area of 10 hectares or less do not need to declare the level of fertiliser used. Such farmers account for more than 70 % of all farmers and use 14 % of the agricultural area. As another example, livestock farmers who keep fewer than 10 livestock units in one place do not need to have a manure storage of the relevant capacity. Such farmers constitute 87 % of all livestock farmers and keep 28 % of all livestock units in Lithuania. The Nitrates Directive aims at protecting the environment and citizens’ health. While the use of thresholds may reduce red tape for smaller farmers, excluding farmers from cross-compliance checks is not only irregular, but may also adversely affect the achievement of the objectives of the Nitrates Directive. Example of insufficient cross-compliance checks The Commission takes note of the Court’s finding. Although the Commission’s 2014 audit in Lithuania only covered animal-related SMRs, similar deficiencies were found in terms of sampling. Those deficiencies are being followed up in a conformity clearance procedure. DG MARE and Member States systems related to the regularity of fisheries transactions Managing authorities make payments to the beneficiaries, and declare the expenditure to the Commission, which then reimburses the Member States. The Commission uses the annual control reports and the annual opinion of the audit authorities to assess whether the Member States have set up management and control systems that comply with the legal requirements. The Commission also carries out its own audits to verify that the systems function effectively. Our examination of systems in this area focused on audits carried out by the Commission (DG MARE). We carried out a desk review of a sample of five of the Commission's conformity audits. The Commission reported significant weaknesses in the systems of two (Italy and Romania) out of the five Member States concerned. We reviewed the Commission’s audit work and its compliance with the international auditing standards. The five audits which we reviewed were not fully compliant with the standards, and we identified weaknesses in audit supervision and audit documentation. The Commission recognised that improvements were required in the areas of audit supervision and documentation, and strengthened its procedures accordingly. The files audited predate these changes. We also re-performed one Commission audit in Romania. We identified deficiencies in the managing authority’s administrative checks, in particular relating to procurement procedures (see Figure 7.11). Procurement deficiencies in Romania We examined a sample of seven procurement procedures, five of which were affected by errors (two public and three private). The three private cases were not detected by the Commission. We also identified deficiencies relating to the project selection procedure in two out of six cases reviewed. Procurement deficiencies in Romania The standard checklists used by the Commission focus on compliance with public, rather than private, procurement rules. The seven procedures referred to concern both private procurement cases, on the basis of instructions drawn up by the Romanian authorities (five cases) and public procurement cases (two cases relating to technical assistance) where the Commission was already aware of the weaknesses. Furthermore, the Commission had previously identified that there were weaknesses in project selection. On the basis of the 24 audits which it carried out in 2014 and 2015 and which were finalised at the time of our audit, DG MARE proposed two financial corrections. Our review of a sample of five of these audits revealed that in two cases (Italy and Romania) the financial corrections did not address all the deficiencies affecting the managing authorities’ internal control systems. The Commission proposes financial corrections where it identifies significant deficiencies. It considers that appropriate levels of correction were applied for both Italy and Romania based on its own analysis. However, for Romania, it will assess the need for additional corrections based on the Court’s findings. DG AGRI’s annual activity report In its 2015 AAR DG AGRI assesses the expenditure at risk for the main spending areas. The starting point of this assessment is the levels of error reported by the Member States (for each paying agency) in their control statistics. Subsequently, the Commission adjusts those levels of error using mainly the Commission's and the Court's audits carried out in the last three years as well as the certification bodies’ opinions on legality and regularity for the 2015 financial year. The resulting adjusted error rates (AERs) are then aggregated and used to calculate the total expenditure at risk, i.e. the level of error that the Commission estimates to affect the EU budget. We consider that this is generally a valid approach which can provide sufficient basis for reservations at the level of individual paying agencies. However, we highlight the following issues. For the first time in 2015, certification bodies were required to ascertain, on the basis of a representative sample, the legality and regularity of the expenditure for which reimbursement has been requested by the paying agencies from the Commission. The certification bodies thus represent a key element in the Commission assurance model for the 2014-2020 period. In 2015, DG AGRI could use the opinions of the certification bodies on the legality and regularity only to a very limited extent due to significant weaknesses in methodology and implementation (such as delays in appointing certification bodies in Member States, inadequate audit strategies, too small samples being drawn for substantive testing and insufficient technical skills and legal expertise to detect deficiencies in paying agencies’ administrative and on-the-spot checks). The Commission considers that having an opinion on legality and regularity by an independent, professional audit body at national level will, over time, result in an improvement in the management and control systems of the Member States and a greater level of assurance on the error rates reported by the Member States. As financial year 2015 was the first year of implementation of this new approach, the full benefits of the new work of the certification bodies could not yet be reaped. Nevertheless, a number of encouraging messages and lessons for the future can be taken from the 2015 exercise. The Commission made adjustments to determine the AER on the part of expenditure for which its own or our audit results were available. However, DG AGRI did not disclose in its AAR which part of the expenditure was not adjusted because no such audit results were available. In such cases, the Commission places reliance on Member States control statistics which may result in an underestimated AER. The Commission examines the situation for each paying agency not only based on its own and the Court’s audit findings over the three previous years, but also on the basis of the opinions received for the financial year concerned from the certification body for each paying agency. There are therefore audit results available in all cases. For the first time in 2015, certification bodies were requested to provide an opinion on the legality and regularity of underlying transactions, which provided even more details and information relevant for deciding on adjusting the reported error rates. THE COURT’S OBSERVATIONS Figure 7.12 gives the key data used to assess DG AGRI’s 2015 AAR. Source: European Court of Auditors based on DG AGRI annual activity reports. Main spending area AAR (financial year) 2015 expenditure (million euro) Average level of error reported by the Member States Aggregated adjusted error rate Corrective capacity Average financial corrections over the last three years Average recoveries over the last three years Total (million euro) (million euro) EAGF 2014 44137,85 0,55 % 2,61 % 536,44 1,22 % 117,83 0,27 % 1,49 % 2015 44834,52 0,68 % 1,47 % 665,36 1,48 % 107,41 0,24 % 1,72 % EAFRD 2014 11186,00 1,52 % 5,09 % 113,58 1,02 % 95,64 0,85 % 1,87 % 2015 9843,14 1,78 % 4,99 % 177,99 1,81 % 114,95 1,17 % 2,98 % THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES For EAGF, in spite of an increase in the average level of error reported by the Member States from 0,55 % in 2014 to 0,68 % in 2015, the Commission concludes that the adjusted error rate has fallen from 2,61 % in 2014 to 1,47 % in 2015. The Commission explained that this reduction of the adjusted error rate is largely caused by: (15) improvements in the Member States’ control systems as a result of remedial actions implemented as from the 2014 claim year; (16) more up-to-date information available to the Commission allowing a reduction of previously calculated adjustments. Txt. 9. The Commission considers that an increase in the average level of the reported errors proves that the Member States’ management and control system work better at detecting and reporting errors. It follows that the reported error rate is considered more reliable and therefore fewer adjustments are needed. The factors mentioned by the Court have positively impacted the evolution of the adjusted error rate. For EAGF, the AAR reports that the aggregated adjusted error rate has fallen from 2,61 % in 2014 to 1,47 % in 2015. Our audit results show that the estimated level of error remained stable, when excluding quantification of cross-compliance errors (see Annex 7.1). For EAFRD, the Commission almost tripled the average level of error reported by the Member States of 1,78 % to calculate the adjusted error rate of 4,99 %. The Commission has adjusted the error rates reported by 56 paying agencies out of 72 declaring EAFRD expenditure. We consider the adjusted error rate of 4,99 % to be broadly in line with our estimated level of error. The Commission has in recent years increased the number of its conformity audits of EAFRD expenditure and also launched horizontal audits targeting measures/paying agencies with similar risks. An example is an audit of public procurement which contributed to overall reservations being made for Germany and Spain. The Commission defines corrective capacity as a mechanism for correcting the errors made and protecting the EU budget. The annual increase in the calculated corrective capacity for EAGF (to 1,72 %) and for EAFRD (to 2,98 %) is the result of recent efforts by the Commission’s to reduce the number of old open conformity audits (see paragraph 7.34). This affected the overall amount of the financial corrections adopted, especially in 2015. The corrective capacity is an estimate of the amounts of future corrections calculated as a historical average of the net financial corrections as explained in the Directorate-General for Agriculture and Rural Development’s 2015 annual activity report. The Commission corrective capacity is calculated on the basis of a three year average of the financial corrections received and has no direct link with the 2015 expenditure which will be corrected in the years to come. This explains this year’s anomaly that the Commission reports its corrective capacity for EAGF expenditure to be higher than the adjusted error rate. For EAFRD, there is no assurance that the remaining financial risk will be below materiality by the end of the programming period. The Commission confirms this in its 2015 AAR. The corrective capacity and the adjusted error rate are two estimates which provide together an indicator of the assurance on the effectiveness of the multiannual control system in protecting the EU budget. In its 2015 annual activity report, the Director-General for Agriculture and Rural Development gives assurance for the 2015 CAP expenditure as a whole, taking into account both the adjusted error rate and the corrective capacity. DG MARE’s annual activity report We examined DG MARE’s 2015 annual activity report, focusing on whether all substantial information was disclosed, whether the information provided was in line with our own findings, and whether the estimations used by the Commission to calculate the error rates were reasonable. DG MARE sent out three pre-closure letters for three operational programs proposing to recover an amount of 53,6 million euro. This material amount is not disclosed in the AAR. The AAR includes reservations for five Member States, including Romania, which is in line with our audit findings. Our audit did not call into question the calculation of the error rates by DG MARE. This amount was not disclosed as the Commission does not consider the amount to be at risk. A recovery order has been issued for one of the three programmes, for the amount anticipated. For the remaining two programmes, when the annual activity report was drawn up the Member State and the Directorate General for Maritime Affaires and Fisheries already largely agreed on the amounts to be recovered by the Commission. Any adjustments to this amount (and the associated financial risk) will therefore be minimal. Conclusion and recommendations The conclusion for 2015 Overall evidence indicates that spending on Natural resources and on both areas subject to a specific assessment (see paragraphs 7.13 and 7.21) is affected by a material level of error. For this MFF heading area, testing of transactions indicates that the estimated level of error present in the population is 2,9 % (see Annex 7.1). The Commission takes note of the most likely error estimated by the Court which is broadly the same as last year excluding cross-compliance. Recommendations Annex 7.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, the Court presented 20 recommendations. The Commission and the Member States fully implemented four recommendations, while five were implemented in most respects and 10 in some respects. The Commission takes note of the latest assessment made by the Court of the progress in addressing recommendations made in previous annual reports. The Commission will continue working with the Member States to further improve the way they follow up the Court’s recommendations. Following this review and the findings and conclusions for 2015, we recommend that the Commission, for EAGF: (17) Recommendation 1: continue its efforts to follow up on cases where national legislation is not compliant with EU legislation, including all legal means at its disposal, in particular suspension of payments (see paragraph 7.20 and Figure 7.5); Txt. 10. The Commission accepts the recommendation. The Commission considers that it is being implemented by means of its ongoing audit work and the provision of guidance when required and takes the necessary steps in the context of shared management to assist the Member State audit bodies to discharge their tasks effectively. In addition, shortcomings in Member States management and control systems are addressed through targeted and comprehensive action plans where necessary. When the Commission detects such problems during the course of its monitoring or audits, they request the Member State to take remedial actions. Where the problem is particularly acute, the Member State is required to implement a remedial action plan which is closely followed by the services. So far such plans have been found to be very effective. When the implementation of the action plan is not deemed to be effective, payments can be reduced or suspended in order to protect the EU budget. (18) Recommendation 2: annually monitor the results of the LPIS quality assessments performed by Member States, and check that all Member States with negative assessments actually take the necessary remedial action (see Figure 7.9); Txt. 11. The Commission accepts this recommendation. The results of the LPIS quality assessment (QA) are monitored on an annual basis. This yearly exercise is a quality control process whose ultimate objective is to improve the quality and reliability of the LPIS, by analysing detected technical weaknesses and to taking remedial actions to address them. for rural development: (19) Recommendation 3: ensure that all Member States’ action plans addressing errors in rural development include effective actions on public procurement (see paragraph 7.45); Txt. 12. The Commission accepts this recommendation. The Commission is fully engaged in providing targeted and effective guidance on how to prevent irregularities in public procurement procedures: Public Procurement — a study on administrative capacity of the EU (2016), including country-specific recommendations, and Guidance for practitioners on the avoidance of the most common errors in public procurement of projects funded by the European Structural and Investment Funds (2015). In December 2015 the Commission has underlined that all Member States need to pay particular attention to public procurement rules, and take corrective action where relevant. The Commission is working intensively with the Member States in order to identify and remedy the causes of errors in the implementation of the policy. This work will continue with the aim to further address identified shortcomings, including requesting action plans. The Commission insists on the completeness and effectiveness of these action plans. Member States are requested to take corrective actions when weaknesses are identified in the context of conformity audits, audit by the ECA or findings of the certification body or paying agency. Finally, net financial corrections are applied where necessary. for EAGF and rural development: (20) Recommendation 4: monitor and actively support certification bodies in improving their work and methodology on the legality and regularity of expenditure so that they provide reliable information for the Commission’s estimation of the adjusted error rate (see paragraph 7.54); Txt. 13. The Commission accepts this recommendation. The Commission will continue to support the Member States to improve and consolidate the work done by the certification bodies, to ensure that, in the coming years, increased assurance can be drawn from their work. (21) Recommendation 5: update DG AGRI’s audit manual by including detailed audit procedures and documentation requirements for the verification of the data supplied by Member States and used for calculating financial corrections (see paragraph 7.36); Txt. 14. The Commission accepts this recommendation. The Commission will when updating its audit manual address this issue building on the experience gained. for fisheries: (22) Recommendation 6: improve the compliance of DG MARE’s conformity audits with international auditing standards (see paragraph 7.50). Txt. 15. The Commission accepts this recommendation. It should be noted that the procedural improvements that took place after the period covered by the Court’s audit have already addressed many of the issues highlighted here (in particular as regards audit evidence and audit trail). Furthermore, a quality control procedure is currently being activated. PART 2 — RURAL DEVELOPMENT PERFORMANCE-RELATED ISSUES Assessment of projects’ performance Last year, we reported our work on performance-related issues for selected rural development transactions. This year, we continued this work. Of the 159 rural development transactions tested for regularity, we assessed performance related issues for 59 investment projects in tangible assets. Of the 59 investment projects 34 were completed at the time of our audit visit. For these completed investment projects we checked whether the investment had been carried out as planned. This criterion was fulfilled for most projects (95 %). For the 59 investment projects we found the following main performance issues: (23) insufficient evidence that costs are reasonable; (24) deficiencies in targeting measures and selecting projects, including weak links to Europe 2020 objectives. Txt. 16. Insufficient evidence that costs are reasonable For 33 of the 59 investment projects (56 %), we found sufficient evidence that project costs were reasonable. However, for 26 investment projects (44 %), this was not the case. Last year the respective figures were 69 % and 31 %. For both years, the projects examined were implemented in the 2007–2013 programming period and under the same rules. Figure 7.13 gives two examples of investment projects for the same rural development measure, one where we found that costs were reasonable, and one where this was not the case. In 2014 we published a special report containing a detailed examination of this topic. The assessment of reasonableness of costs is the responsibility of the Member States (see article 48 of Regulation (EU) No 809/2014). The Commission protects the EU budget when weaknesses are identified through conformity procedures and issues guidance on how to improve the systems in place (see Guidance on rural development controls and sanctions, including a checklist for the assessment of reasonableness of costs). Furthermore, training and sharing of good practices are organised by the European Network of Rural Development. Finally, the Commission is actively promoting the use of simplified cost options in the Member States. Examples of checking reasonableness of costs for the modernisation of agricultural holdings In Greece, a beneficiary bought a tractor partially financed by rural development funds. The beneficiary had to use a predetermined formula to calculate the appropriate horsepower of the tractor, taking into consideration the cultivated area of the land. The beneficiary provided an offer which the paying agency verified against a reference price which was taken from a specialist magazine about farm machinery. The paying agency concluded that the price for the specific model should be 10 % lower than the price proposed. As a result, the beneficiary bought a cheaper tractor. In Slovakia, we audited a project for the construction of a storage facility. The paying agency did not check if the project costs were reasonable. However, we obtained conclusive evidence that the price of one of the main construction materials (concrete) was six times higher than the normal market price. Overall, the cost of the project should have been more than 50 % lower. Examples of checking reasonableness of costs for the modernisation of agricultural holdings For the Slovakian case the Commission notes that the project was implemented. Out of the 59 investment projects, 50 (85 %) were in line with the needs for both quantity and quality. For example, we visited three projects that relate to the construction or the extension of commune centres in small towns or villages under the measures village renewal and development and implementing local development strategies. The investments were fully utilized at the time: all the rooms were in use and activities for the local community were scheduled for weeks ahead. However, for nine projects (15 %) the output of the project was larger than actually needed. One example of such a case is the construction of a warehouse which was several times bigger than needed and therefore remained for a large part unused. Another example of a project which included non-essential items is given in Figure 7.14. Example of a project which included non-essential items In Italy (Campania), a municipality renewed a 1-kilometre footpath in a mountainous area for a cost of 441000 euro. During the procurement procedure, the municipality asked interested companies for proposals to make the investment more functional and environmentally friendly. The proposal by the company selected included additional items costing 80000 euro, over and above the cost of the work on the footpath. These additional items included a mountain bike worth 4000 euro, a panoramic spyglass costing 3500 euro, and a 10000 euro donation to a local church. Instead of restricting the contract to the needs of renovating the footpath, additional items were added which led to higher costs. Example of a project which included non-essential items The project in question, supported under the forestry measure 227 Non-productive investments, is under investigation by the Region. The Regional Managing Authority has not paid the final balance of the project yet and intends to make a reduction as a result of the control carried out. Deficiencies in targeting the measures and selecting projects We tested whether support was targeted at clearly defined objectives reflecting structural and territorial needs and structural disadvantages. This criterion was not fulfilled for 22 % of the investment projects examined. Last year we reported a very similar percentage (21 %). Examples of good and bad practice are given in Figure 7.15. We have reported on the lack of adequate targeting for rural development investments in our special reports. For the 2014-2020 period, selection criteria have to be established for all measures but the area- and animal-related ones. The principles of selection criteria, in line with strategy and objectives of the policy, must be set in the programme. Applications will then have to be assessed and ranked against specific selection criteria on which the Monitoring Committee should be consulted. Minimum thresholds should be established for the application of selection criteria to avoid financing poor projects. Examples of targeting and selection In Italy (Campania), the paying agency opened a call for proposals with a budget of 25 million euro under the measure modernisation of agricultural holdings. The paying agency applied different selection criteria as regards the beneficiary (e.g. criteria on young farmers, gender balance), the holding (e.g. criteria on holdings situated in less favoured areas or producing quality products) and the project (e.g. criteria on synergies with other measures, innovation etc.). The Italian authorities also set a minimum threshold in advance in order for a project to be selected. In this case, the Italian authorities prioritised the principles of sound financial management above spending. In Hungary, the objectives of the measure modernisation of agricultural holdings include three broad areas (upgrading, diversification and compliance with standards). The audited sub-measure was targeted towards animal holdings with a certain minimum size. There was no other prioritization or geographical targeting of the aid. The rural development programme split the available budget into amounts that become available periodically to applicants. Although the Hungarian authorities defined selection criteria, they did not use them. The call for proposals for the audited project opened on 16 December 2013, for a budget of 10 million euro. The Hungarian authorities funded projects on a first come, first served basis. The last project accepted was submitted on 19 December 2013, only three days after the opening of the call. In this case, the selection process took no account of the projects’ potential impact. The approach observed in Hungary allows the paying agency to fund projects with lower value-added, for the simple reason that they are submitted earlier than other projects with higher added value. Examples of targeting and selection The added-value of projects is taken into account as part of the eligibility checks. At the very end of the programming period 2007-2013, in December 2013, Hungarian authorities did not use the same selection system as during the previous years. The objective was to support investments that can be still finished before the end of 2015. For the 59 investment projects examined, we assessed whether Europe 2020 targets such as innovation or renewable energy were translated into operational elements for targeting and selection procedures. We found this was the case in only 51 % of the projects reviewed, despite the fact that almost all of them (57 out of 59) had been approved after the adoption of Europe 2020 targets. Examples of good and bad practice are given in Figure 7.16. Projects selected by national authorities during the 2007-2013 programming period should not be evaluated against Europe 2020 objectives as those had not been defined when the legal framework for the programming period was adopted (in 2005). In that light, the Commission considers it a positive result that for 51 % of the projects reviewed Europe 2020 targets were reflected. The Commission considers that rural development priorities as set out for 2014-2020 are a translation of thematic objectives of Europe 2020 into the specific context of the rural development policy. EAFRD legal framework establishes a clear link: the contribution to Europe 2020 shall be pursued through six priorities subdivided into focus areas. This requirement is fully reflected in the structure of the EAFRD programmes. The progress towards targets set for the priorities/focus areas is monitored in the Annual Implementation Reports. In order to ensure consistency with the thematic objectives, correspondence tables have been annexed to the guidance documents. Examples of links to EU priorities In Italy (Campania), we audited a project with a value of almost 100000 euro under the measure non-productive investments. The project consisted in rebuilding dry stone walls with stones recovered from a previously existing wall of a similar size, aimed at reconstructing elements of the rural landscape. For this case, we did not identify a link to any of the objectives of the EU 2020 Strategy. In this region we also audited a project which was selected on the basis of the procedure described in Figure 7.15. The project included investments in a new irrigation system, an area for rain water collection and photovoltaic panels. The project contributes to adaptation to climate change and energy efficiency, which is one of the five headline targets of the EU 2020 Strategy. Examples of links to EU priorities Rebuilding dry stone walls is an action covered, under the 2007-2013 rural development programming period, by measure 216, with the objectives of land protection and conservation of biodiversity. The investment is relevant in relation to those objectives and in line with the EU biodiversity strategy to 2020 (COM(2011) 244 final). Biodiversity, renewable energy and water management are closely linked to each other and can all contribute to the climate change challenges. Conclusion Our examination of performance aspects for rural development investment projects showed that most investments were carried out as planned. However, we found deficiencies in targeting and a lack of evidence that costs were reasonable. In the projects examined, Member States did not sufficiently take EU priorities into account. For these issues we did not notice an improvement compared to last year. The Commission has identified similar weaknesses in the Member States’ assessment of reasonableness of costs in the framework of audit conformity procedures, applying where relevant financial corrections to protect EU budget. Furthermore, the Commission has published guidance on controls and penalties under rural development, including the checklist to assess reasonableness of costs developed by the Court of Auditors in the special report No 22/2014. Finally, the Commission is actively advocating the use of simplified cost options under investment measures, which would contribute to the avoidance of such type of irregularities. As regards targeting, the Commission notes that this is ensured through the intervention logic of the policy and the programmes: EAFRD-supported operations (projects) must comply with objectives of the measure programmed in the RDP, applying adequate selection criteria. The measures contribute to the rural development priorities, which are consistent and reflect EU2020 strategy for smart, sustainable and inclusive growth (see Article 5 of Regulation (EU) No 1305/2013). The common monitoring and evaluation system facilitates tracking of results and evaluates impacts in overall strategies (CAP and EU2020).

THE COURT’S OBSERVATIONS INTRODUCTION This chapter presents our findings for Natural resources. The chapter presents separately the results of testing of the European Agricultural Guarantee Fund (EAGF) and for other spending (Rural development, environment, climate action and fisheries) under the multi-annual financial framework heading (MFF). Key information on the composition of this MFF heading is provided in Figure 7.1. MFF heading 2 Natural resources — Key information 2015 (billion euro) Source: 2015 consolidated accounts of the European Union. Total payments for the year 58,0 - advances 2,4 + clearings of advances 2,9 Audited population, total 58,5 THE COURT’S OBSERVATIONS In addition to our findings with regard to regularity for the specific assessments, this chapter presents, in a separate section, the results of some performance-related issues for a sample of rural development projects. Brief description of Natural resources The common agricultural policy (CAP) is the basis for EU spending on agriculture. The Commission, in particular the Directorate-General for Agriculture and Rural Development (DG AGRI), and paying agencies in the Member States share the management of implementation of the CAP (see paragraph 7.31). They pay beneficiaries through two funds: the EAGF, which fully finances EU direct aid and market measures, and the European Agricultural Fund for Rural Development (EAFRD), which co-finances rural development programmes together with the Member States. The main measures financed by the EAGF are:

The direct aid Single Payment Scheme (SPS), 29,3 billion euro in 2015, and Single Area Payment Scheme (SAPS), 7,8 billion euro in 2015. These area-related schemes provide decoupled income support.

Other direct aid schemes, 5,1 billion euro in 2015, and interventions in agricultural markets, 2,7 billion euro in 2015.

The EAFRD co-finances the rural development expenditure disbursed through Member States' rural development programmes. The expenditure covers both area-related measures and non-area-related measures. In 2015, the EAFRD accounted for 11,8 billion euro of expenditure. The main risks to regularity are that:

direct aid is paid for ineligible land or to ineligible beneficiaries;

aid for interventions in agricultural markets is granted for ineligible costs or to ineligible applicants;

rural development expenditure is ineligible, due to non-compliance with often complex rules (including procurement) and eligibility conditions, in particular for investment measures.

This chapter also covers the EU environmental policy, climate action, and the common fisheries policy. The main risk to regularity in these areas is that aid is granted for ineligible or overstated costs. Audit scope and approach Annex 1.1 sets out the key elements of our audit approach and methodology. For the audit of Natural resources, the following specific points should be noted:

(a) We examined a sample of 180 transactions for EAGF and 179 transactions for Rural development, environment, climate action and fisheries as defined in Annex 1.1, paragraph 7. Each sample is designed to be representative of the range of spending under the MFF heading for each of the two specific assessments. In 2015 the EAGF sample consisted of transactions from 18 Member States. For the second specific assessment the sample consisted of 159 transactions for rural development and 20 other transactions including environment, climate action and fisheries from 17 Member States.

(b) We assessed selected internal control systems through a desk review of 24 Commission audits covering the EAGF and the EAFRD. We complemented this work by re-performing six of the Commission’s audits in the Member States concerned. We performed additional audit work in these six Member States for a limited number of risk-prone areas relating to the EAGF and the EAFRD, as well as for cross-compliance. We also reviewed five Commission audits covering the common fisheries policy, and re-performed one.

THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES

(c) We assessed the annual activity reports of DG AGRI and DG MARE.

For both the EAGF and the EAFRD, 2015 was the last year when most payments were made to final beneficiaries on the basis of the rules prevailing before the 2013 CAP reform. PART 1 — OUR ASSESSMENT ON REGULARITY Regularity of transactions Annex 7.1 contains a summary of the results of transaction testing for Natural resources as a whole as well as for each of the two specific assessments (EAGF and Rural development, environment, climate action and fisheries). Of the 359 transactions examined, 121 (34 %) were affected by error. On the basis of the 87 errors we have quantified, the estimated level of error for Natural resources is 2,9 %. The Commission takes note of the most likely error estimated by the Court which is broadly the same as last year excluding cross-compliance. The Commission is of the view that net financial corrections resulting from multiannual conformity procedures, as well as recoveries from beneficiaries reimbursed to the EU budget, represent a corrective capacity that has to be taken into account in a comprehensive assessment of the overall system of internal control. The Commission also notes that, as reported in the 2015 annual activity report of the Directorate General for Agriculture and Rural Development, its corrective capacity amounted in 2015 to 1065,710 million euro (1,87 % of the total CAP expenditure). Figure 7.2 presents the breakdown of the different types of errors to our overall estimated level of error for 2015. Breakdown of the estimated level of error — Natural resources Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The nature and pattern of errors is significantly different between the two specific assessments, as set out in the following paragraphs. The Commission notes that there are significant differences in the design, scope and objectives of both pillars of the CAP which explain these differences in the pattern of CAP-related errors. EAGF — Market and direct support Annex 7.1 contains a summary of the results of transaction testing. For the EAGF, of the 180 transactions that we examined, 52 (29 %) were affected by error. On the basis of the 47 errors we have quantified, the estimated level of error is 2,2 %. The Commission takes note of the most likely error estimated by the Court which is broadly the same as last year excluding cross-compliance. The Commission also notes that the frequency of errors has steadily decreased over the last three years. Figure 7.3 presents the breakdown of the different types of errors to our estimated level of error for 2015. Breakdown of the estimated level of error — EAGF Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES In four cases of quantifiable errors made by final beneficiaries, national authorities had sufficient information to prevent, or detect and correct the errors before declaring the expenditure to the Commission. If all this information had been used to prevent, or detect and correct errors, the estimated level of error for this specific assessment would have been 0,3 percentage points lower. In addition, we found that for another four cases, the error that we detected had been made by the national authorities. These errors contributed 0,6 percentage points to the estimated level of error. We give an example of errors made by the national authorities in Figure 7.5. The Commission welcomes the Court’s assessment on Member States’ role and is of the view that Member States should have done more to reduce errors. The Commission continues to work with Member States to reduce errors. Most of the quantifiable errors related to overstatements of the eligible area of land (see Figure 7.3). In 2015, we found such errors in 12 out of the 18 Member States visited. Around half of these quantifiable errors were smaller than 2 % (see paragraph 7.18) and thus had only a limited impact on the overall estimated level of error. A further quarter of these errors were between 2 % and 5 %. The Commission takes note that most of the over-declaration errors detected by the Court were relatively small errors and welcomes the Court’s assessment on the improvement of the reliability of LPIS data. The Commission considers that minor errors are impossible to avoid at a reasonable cost. The Commission encourages Member States to update their land parcel identification system (LPIS) on a regular basis to ensure detecting large over-declarations. The Commission considers the fact that the large majority of the errors reported by the Court were under 5 % to be proof of the good functioning of the integrated administrative control system (IACS) of which LPIS is a fundamental component. As land changes over time, differences between LPIS and the reality on the ground may occur, however regular updates of the system limit the risk. To promote frequent updates of the LPIS, the Commission has provided for, as a pre-condition to reduce on-the-spot checks from 5 % to 1 % for some direct payments schemes, an LPIS update at least every three years with ortho-images aged less than 15 months (the update covering per year should be at least 25 % of the eligible hectares). A quality assessment of LPIS was also introduced in 2010 and maintained in the present period (Regulation (EU) No 640/2014). This yearly exercise done by Member States’ administrations and monitored by the Commission is a quality control process whose ultimate objective is to improve the quality and reliability of the LPIS, by analysing detected technical weaknesses and by taking remedial actions to address them. The annual on-the-spot checks are still a way, on a sample basis, to achieve an accurate and updated LPIS. In the Member States, the key control tool for area aid is the Land Parcel Identification System (LPIS), which is a database containing eligibility data for all agricultural parcels. It contains computerised graphical information, including ortho-images (aerial photographs) which Member States should update at least every five years. Provided the graphical information is properly updated and analysed, the LPIS makes a significant contribution to preventing and detecting area-related errors in aid applications. Thanks to action plans set up by the Commission and the Member States, the reliability of data in the LPIS has been constantly improving over recent years. Nevertheless, we still detect area-related errors. This is because administrative checks on the eligibility of land are based on the same LPIS data which farmers frequently copy into their aid applications. Therefore, if the LPIS data is incorrect, such checks cannot normally detect eligibility errors in the claim. The technical standard required for ortho-images does not usually allow the farmers and the paying agencies to detect small errors (smaller than 2 %, see paragraph 7.16), which are, to a large extent, impossible to avoid at a reasonable cost. However, the larger area-related errors were, in general, detectable. In such cases, the eligible area recorded in the LPIS database and declared by the farmer was substantially incorrect, mainly because:

the paying agencies did not reduce payments sufficiently, or at all, for ineligible vegetation and features visible on the ortho-images of the land parcel concerned when assessing the eligible area for the purpose of updating the LPIS, or

national definitions of eligible permanent pasture did not comply with the EU definition.

Whenever there is a case of national definition for permanent pasture that did not comply with the EU definition, specific corrective actions including, if necessary, action plans are set up to address the issue. To protect the EU budget, the Commission also applies reductions and suspensions of payments and net financial corrections in the framework of conformity clearance procedures whenever necessary. We present an example of incorrect eligibility data in the LPIS in Figure 7.4. Example of irregular payments due to incorrect eligibility data in the LPIS Permanent pasture is defined as land used to grow grasses or other herbaceous forage. In our 2014 annual report we reported that in Spain, despite considerable remedial action, important weaknesses in the LPIS relating to the eligibility of permanent pasture still persisted, and that in 2014 and 2015 further corrective action was initiated. We inspected a parcel of 48 hectares in Spain (Madrid), recorded in the LPIS as fully eligible permanent pasture at the time the farmer made the claim. The ortho-image showed however the presence of dense bushes and trees. In his 2014 aid application, the farmer had declared the parcel on the basis of the eligible area recorded in the LPIS and he was paid SPS aid for the entire parcel. We found the parcel to be ineligible for SPS aid. In 2015 the Spanish authorities carried out a review of their LPIS on the basis of new ortho-images and using new technology, and recorded the parcel concerned as ineligible. We also found that area aid was claimed and paid on the basis of incorrect eligibility information in the LPIS, which was detectable on the ortho-images, in Greece, Spain (Aragon, Castilla-La Mancha, Castilla y León and Extremadura), Italy (Calabria) and the UK (England). Example of irregular payments due to incorrect eligibility data in the LPIS Issues around permanent pasture have been identified by the Commission before 2014 and action plans have been implemented in the Member States concerned. In Spain, the national authorities carried out full re-measurement of all the parcels declared by the beneficiaries audited by the Court and launched recovery procedures where necessary. Furthermore, the fact that the parcel under consideration was recorded as ineligible in 2015 proves that the procedure to update LPIS in Spain works. Audits carried out by the Commission services in previous years had identified deficiencies linked to the eligibility of permanent pasture in Greece and Spain and ensured that they were addressed through specific action plans which led to improvements (see Court’s 2014 annual report, box 7.8). In both Greece and Spain, the area recorded as eligible in LPIS was significantly reduced: in Greece the eligible permanent pasture area was reduced from 3,6 million hectares in 2012 to 1,5 million hectares in claim year 2014 and in Spain, the eligible permanent pasture area was reduced from 18,5 million hectares in 2013 to 6,4 million hectares in 2016. For the UK (England) the remedial action plan is ongoing and is being closely monitored by the Commission. Furthermore, conformity clearance procedures ensure that the risk to the EU budget is adequately covered by net financial corrections. We present an example of national eligibility criteria not complying with EU legislation in Figure 7.5. Example of national eligibility criteria not complying with EU legislation In our 2013 annual report we reported that French authorities grant aid for grazable heathland, although these areas comprise a mixture of eligible herbaceous vegetation and ineligible bushes, dense woods, etc. In 2015 we found six such parcels to be fully or partially ineligible because they contained dense trees and bushes, which the French authorities had assessed as fully eligible. Example of national eligibility criteria not complying with EU legislation Audits carried out by the Commission prior to 2013 had already identified similar deficiencies in France. The weaknesses are being followed up under a specific action plan established by the Member State and closely monitored by the Commission. Any such weaknesses that are found are followed up through conformity clearance procedures which ensure that the risk to the EU budget is adequately covered by net financial corrections. In parallel, the Commission remains vigilant that the action plan is adhered to in order to ensure improvement. In light of certain delays incurred further measures pursuant to article 41(2) of Regulation (EU) No 1306/2013 were initiated as of 2015. Rural development, environment, climate action and fisheries Annex 7.1 contains a summary of the results of transaction testing. For Rural development, environment, climate action and fisheries, of the 179 transactions that we examined, 69 (39 %) were affected by error. On the basis of the 40 errors we have quantified, the estimated level of error is 5,3 %. The Commission takes note of the most likely error estimated by the Court, which is lower than that observed last year. Figure 7.6 presents the breakdown of the different types of errors to our estimated level of error for 2015. Breakdown of the estimated level of error — Rural development, environment, climate action and fisheries Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES For rural development, of the 159 transactions we examined, 78 were area-related and 81 non-area-related. Of these 159 transactions, 64 (40 %) were affected by error, of which 38 were quantifiable. As regards environment, climate action and fisheries, of 19 transactions sampled, 5 (26 %) were affected by error, of which 2 were quantifiable. For the one remaining transaction, we found no error. For rural development, the frequency of errors (both quantified and not-quantified) found by the Court has steadily decreased over time. In 13 cases of quantifiable errors made by final beneficiaries, national authorities had sufficient information to prevent, or detect and correct the errors before declaring the expenditure to the Commission. If all this information had been used to prevent, or detect and correct errors, the estimated level of error for this specific assessment would have been 1,7 percentage points lower. In addition, we found that for six cases, the error that we detected had been made by the national authorities. These errors contributed 1,5 percentage points to the estimated level of error. We give an example of an error made by the national authorities in Figure 7.7. The Commission is aware that the national authorities could potentially have detected many of the errors found by the Court. The CAP rules provide the Member States with all necessary instruments to mitigate most of the risk of errors. For the example given in Figure 7.7, the Commission has launched in March 2016 a conformity clearance procedure. All necessary actions will be taken to protect the EU budget. The types of errors are similar to those we found in previous years and to those identified in our special report analysing the causes of errors in rural development spending. As in previous years, non-area related measures were the major component (82 %) of the estimated level of error reported in paragraph 7.21. The Commission is aware of the root causes of errors and is working with the Member States to remedy the situations by offering guidance or requesting action plans to be implemented. For instance, the Commission knows that non-area-related measures are more prone to errors. For that reason systemic solutions are promoted, such as simplifying the design of measures or using less complex calculation methods, for example simplified cost options or lump sums. The Commission is of the view that irregularities in the public procurement procedure do not necessarily mean that the overall expenditure was ineligible. In order to assess the financial risk of public procurement errors, the Commission adopted in December 2013 guidelines for determining financial corrections for non-compliance with public procurement rules in projects implemented under the shared management Of the 81 non-area related transactions examined, 15 (19 %) did not comply with the eligibility requirements (an example is given in Figure 7.7). For environment, climate action and fisheries, the 2 quantifiable errors were due to ineligible expenditure. Example of ineligible expenditure or activity The rural development measure animal welfare payments provides support to farmers who voluntarily make animal welfare commitments. Payments are made annually and cover additional costs and income foregone resulting from the commitment made, such as increased space for animals. This is an example of using simplified cost options, which may reduce the administrative burden for Member States and beneficiaries. Such options should be based on a sound methodology, in particular an accurate calculation of the simplified cost. In Romania we found ineligible expenditure because the national authorities, when calculating the simplified costs, did not appropriately take into account the number of production cycles. As a result, payments to all beneficiaries were systematically overstated. We checked all types of commitments for the animal welfare payments in Romania, including those outside the sample, in order to determine to what extent the miscalculations affected the entire expenditure for animal welfare in Romania. We found that total ineligible expenditure from the time the measure started (2012) to October 2015 could be up to 152 million euro (out of 450 million euro in total). We also found eligibility errors in Belgium (Wallonia), Germany (Rhineland-Palatinate), Spain (Andalucia, Galicia, Castilla-La Mancha), Greece, Italy (Rural network, Sicily and Campania) and Poland. Example of ineligible expenditure or activity The Commission fully supports Member States in their efforts to use whenever possible simplified costs options as a tool to prevent errors, as recognised by the Court in its special report No 23/2014 (see paragraph 86). As regards the Romanian cases, the Commission agrees that there was an error in the calculation of the simplified cost used for the payments. Following the Court’s findings, the Commission launched a conformity clearance procedure in 2016 to establish precisely the amount at risk. This may result in the application of a financial correction. For 27 transactions examined, the beneficiaries had to comply with public procurement rules. These rules ensure purchase of goods and services on the most favourable terms, while guaranteeing equal access to public contracts and compliance with the principles of transparency and non-discrimination. In six cases one or more of these rules had been breached. We present an example of such an error in Figure 7.8. The Commission has identified public procurement as an important source of errors and intensified its audit work to ensure appropriate protection of the EU budget. Compliance with public procurement rules is also one of the central elements of the action plans for rural development. In-depth audits of that area were carried out in 2015/2016 in four Member States. The practice of in-depth checks on public procurement rules continues in other Member States in 2016. In order to assess the financial risk of public procurement errors, the Commission adopted in December 2013 guidelines for determining financial corrections for non-compliance with public procurement rules in projects implemented under the shared management. Example of non-compliance with public procurement rules A commune in Italy (Campania) received support for the construction of a monorail: infrastructure to help farmers transport their olives up and down the hills. The contract was awarded to a private company through a public procurement procedure. In order to assess the various offers, the commune established a set of criteria, with different weightings (measured in points). The contract notice required the bidders to include, as part of their offers, proposals for additional works concerning new roads, sewage and an aqueduct. These additional works accounted for 75 % of the total points on which the commune based the award decision. By comparison, price only accounted for 10 % of the total points. Since the EU aid was granted for the construction of the monorail, the beneficiary breached Italian public procurement law by setting up criteria — the construction of roads, sewage and an aqueduct — which were not related to the subject, nature and characteristics of the contract notice. Such criteria were therefore irregular under Italian public procurement rules. We also found cases of non-compliance with public procurement rules in Germany (Rhineland-Palatinate), Greece and Romania. Example of non-compliance with public procurement rules The Commission understands that the award criteria as set in the call for tender were not fully related to the nature of the works. However, the Commission underlines that the investment was carried out as planned. The Italian authorities reviewed the file and instigated recovery procedure for breaches in the public procurement procedures. The Italian authorities applied a 10 % financial correction. The Commission considers that this corrective action helps protect the EU budget. Reservations were made in the 2015 annual activity report to cover public procurement issues, as recognised by the Court in point 7.60. Our sample included 45 transactions for agri-environment payments, which concerned the use of agricultural production methods compatible with the protection of the environment, landscape and natural resources. We found that in three cases farmers had not met all the conditions for payment. The Commission would like to emphasise that all three beneficiaries had a wide range of different agri-environmental commitments to observe and implemented successfully the majority of them. Only few of the commitments were not respected fully or not properly documented. Of the 78 area-related transactions, we found 15 cases (19 %) where the eligible area of land was overstated. Most of the overstatements we identified were relatively small, and thus had only a limited impact on the overall estimated level of error. See also paragraph 7.18. The Commission takes note that most of the over-declaration errors detected by the Court were relatively small errors and welcomes the Court’s assessment on the improvement of the reliability of LPIS data (see paragraph 7.17). The Commission considers that minor errors are impossible to avoid at a reasonable cost. National authorities should check the reasonableness of costs. In 24 investment projects, the national authorities had not checked this adequately. We reported these cases as errors. However, it is generally not possible to determine the amount of ineligible expenditure, in which case we cannot quantify the error. We include additional information on the reasonableness of costs in the section testing of performance criteria (see paragraph 7.71). See Commission’s reply to point 7.71. Examination of annual activity reports and other elements of internal control systems DG AGRI’s systems related to the regularity of transactions in agriculture The Commission and Member States share the management of almost all agricultural expenditure. Some 80 paying agencies in the Member States disburse the aid to farmers and other beneficiaries; the Commission reimburses the Member States for these costs each month (for the EAGF) or each quarter (for the EAFRD). The Commission carries out conformity audits aimed at assessing whether Member States’ systems ensure that aid payments to beneficiaries are legal and regular. Systems weaknesses detected may lead to financial corrections, imposed on Member States by way of conformity decisions. Financial corrections are often made on a flat-rate basis and cover expenditure relating to several budgetary years. They rarely result in recoveries of payments from final beneficiaries and are thus paid from national budgets. While it is true that the majority of corrections in the past were flat rate, in the new legal framework the Member States are encouraged to provide a calculation for the risk to the Fund. Where undue payments, which can be directly linked to a beneficiary, are identified as a result of the Commission’s conformity clearance procedures, Member States are required to launch the necessary recovery actions against the final beneficiaries. To check that the Commission’s conformity audits complied with international auditing standards and the main regulatory provisions, we carried out desk reviews of a sample of 24 out of 120 conformity audits reported by DG AGRI in 2015. We also re-performed six of these 24 audits. In 2015, the Commission updated its audit procedures following a self-assessment of its audit methodology and quality assurance policy. We reviewed the updated procedures and took them into account for our assessment. The Commission’s conformity audits are overall compliant with the relevant standards. However, when the Commission does not raise any findings, it needs to improve the way it documents the audit work supporting such conclusion. The Commission strives to respect ISSAI when performing its audit. This is one of the reasons that led it to use the services of an external company to support its self-assessment of its audit procedures. The audit reports of each mission are the basis for the findings raised by the Commission. When no findings are raised (e.g. there was no issue to report, the weaknesses detected were not material), this is stated in the audit report. In previous annual reports, we have drawn attention to the problem that conformity procedures were too lengthy, resulting in a significant backlog of open files. In 2015, the Commission significantly reduced the number such old open files: from 192 to 34. Changes in legislation for the 2014-2020 period, which took effect for the CAP on 1 January 2015, were designed to streamline the conformity clearance procedure by introducing legal deadlines. As a result, the Commission now monitors the audit cycle more closely in order to comply with the internal and external deadlines. The Commission follows up closely the closure of backlog audits (old open files): their number has been further reduced from 34 to 20 in June 2016. The Commission’s objective is to close all remaining backlog audits (old open files) by the end of 2016. The new legislation for the 2014-2020 period also placed greater emphasis on the Member State’s role in establishing the amount at risk, which the Commission may use to calculate the financial correction to be imposed. We reviewed whether the Commission correctly validated and used the Member States’ data by examining the Commission guidance and checking six financial corrections adopted or prepared in 2015. As recommended by the Court in previous years, the Commission has developed a framework that is intended to increase the number of calculated financial corrections and thereby apply less flat rate corrections. Taking into account the inherent difficulty in cases where numerous transactions are concerned, appropriate legal provisions have been inserted in the basic act, including the principles of Member States’ responsibility to provide the necessary information and of proportionate effort. The delegated act contains detailed provisions on the calculated and extrapolated financial corrections. The Commission’s detailed guidelines on financial corrections have been updated accordingly, notably regarding the criteria and methodology to be used by the Commission to accept or not the Member States’ calculation or extrapolation of the risk to the EU budget. The Commission considers that the audit manual includes information on the verification of data supplied by Member States. Communication C(2015) 3675 attached to the audit manual sets out clearly the minimum requirements that Member States have to meet when they submit additional data to the Commission for the calculation of the financial corrections and the criteria and methodology to be used by the Commission to accept or not the calculation or extrapolation of the financial correction. It is standard procedure for the Commission to do thorough checks on the data provided by Member States. This is indeed confirmed by the findings of the Court for four of the six financial corrections examined. As regards the two other financial corrections mentioned:

for one, the amount involved was relatively small and additional information was requested from the Member State,

for the 5,8 million euro financial correction, the Commission checked the validity of the data given by the Member State although this process was not properly documented.

The Commission would like to underline that the efforts to check the figures provided by Member States should be proportionate to the potential financial consequences. The Commission can also rely on the work of the certification bodies. The Commission audit manual does not include detailed audit procedures and documentation requirements for the verification of the data supplied by the Member States. For four of the six financial corrections examined, we found that the Commission carried out audit visits to verify the data received from the Member States concerned. For the other two financial corrections the Commission did not provide us with sufficient and appropriate evidence of their checks or verifications. For one of these cases, the financial correction was relatively small. However, for the remaining case, the financial correction was 5,8 million euro. The use of non-verified data creates a financial risk to the EU budget, as it is in the Member States’ interest to reduce financial corrections. Member States’ systems related to the regularity of transactions in agriculture Member States' authorities are responsible for putting in place and operating:

(a) appropriate administrative and control procedures to ensure the accuracy of declarations made by claimants and the fulfilment of eligibility requirements;

(b) on-the-spot checks which, depending on the aid scheme, should cover at least 5 % of all beneficiaries or of the relevant expenditure.

We based our systems examination on audits carried out by the Commission (DG AGRI). Within the sample of 24 conformity audits we reviewed (see paragraph 7.32), the Commission reported significant weaknesses for 17 of the systems concerned. Since we found that the Commission’s work was overall satisfactory (see paragraph 7.33), we can rely on the Commission’s assessment of these systems. The Commission welcomes the Court’s statement that the Commission’s work was overall satisfactory and that it can rely on the Commission’s assessment of Member States’ systems. In the six Member States we visited (see paragraph 7.32), we also performed additional audit work for a limited number of risk-prone areas. Our findings for the EAGF, rural development, and cross-compliance are set out in the following paragraphs. Member States' systems related to regularity of EAGF transactions The Integrated Administration and Control System (IACS) is the main management and control system for direct aid payments. The system consists of databases of farm holdings and aid applications, a Land Parcel Identification System (LPIS; see also paragraphs 7.17 and 7.18), animal databases and a database of entitlements in the Member States implementing the SPS. The paying agencies carry out administrative cross-checks between these databases with the objective of ensuring that payments are made for the correct amount, to eligible beneficiaries for eligible land or animals. The Commission considers that IACS, including LPIS, is a reliable system and that the quality and coverage of IACS is essential for assuring the regularity of direct aid payments and EAFRD expenditure, in particular agri-environmental measures and less-favoured areas. Whenever weaknesses occur, the Member States must put in place corrective actions, including specific action plans, to address them. These are carefully monitored by the Commission. In our 2014 annual report, we stated that IACS makes a significant contribution in preventing and reducing the levels of error in the aid scheme to which it applies. This year, we examined the reliability of selected key control functions in the IACS at the six paying agencies visited (see paragraph 7.32). Without calling into question IACS contribution in preventing and reducing the levels of error, we found that some of the key control functions we examined were affected by weaknesses. These related to the LPIS, the administrative controls, the quality of on-the-spot inspections and the recovery procedures for incorrect payments. We provide an example of the weaknesses we identified in Figure 7.9. Example of the weaknesses found concerning the LPIS For six consecutive years (2010-2015) in the United Kingdom (Northern Ireland) the annual LPIS quality assessment showed that it failed to comply with several quality elements required by EU legislation. Despite the legal obligation to put in place remedial action in such situations, the quality assessment for 2015 showed that there was no consistent improvement: overall, the eligible areas recorded in the Northern Ireland LPIS were overstated by more than 2 %, leading to a risk of irregular payments. We have already reported the weaknesses in the Northern Ireland LPIS in our 2012 annual report (Annex 3.2). Example of the weaknesses found concerning the LPIS From 2013, corrective actions were taken by the Northern Irish authorities following audits carried out by the Commission. The estimation of eligible area in the LPIS quality assessment (QA) was non-conforming for 2010, 2011, 2012 and 2015. The reasons were explained by the UK-NI authorities in the respective annual QA reports. In the years when the 3-year LPIS refresh started (2013, 2014), the number of non-conforming parcels decreased. However, in 2015, the UK-NI authorities detected that there was an increase of the number of parcels with no eligible area which had not been identified through the refresh programme. The Commission continues to closely monitor the situation. Member States' systems related to regularity of transactions in rural development For five out of the six paying agencies visited (see paragraph 7.32), we identified deficiencies in administrative checks related to eligibility conditions, in particular those concerning public procurement. These system weaknesses were very similar to those found and reported in previous years. System failures are an important cause of the errors detected during transaction testing (see paragraph 7.24). The Commission carries out conformity audits in the Member States to verify that the expenditure paid is in compliance with the rules. In these audits, the Commission also found weaknesses in administrative checks related to eligibility conditions and public procurement. Corrective actions have been implemented, including when necessary action plans. In addition reservations were made in the 2015 annual activity report. Financial corrections have also been made in this respect and a number of conformity clearance procedures are currently ongoing and are likely to lead to further financial corrections. A dedicated audit of public procurement was launched in 2015 in 4 Member States. Other Member States will be audited in 2016. For public procurement, we examined a sample of 41 projects (in addition to our transaction testing), of which eight (20 %) were affected by errors. In three out of the eight cases, we identified ineligible expenditure, amounting in total to more than 1,5 million euro of EAFRD aid. In an effort to reduce the level of error in rural development spending, the Commission and the Member States are implementing action plans to identify the causes of error and to implement targeted corrective action to reduce the risk of errors occurring. We reviewed whether, for frequent errors, the action plan of the Member State concerned dealt with the root cause of these errors. We looked at 14 of the most significant quantifiable errors detected in eight Member States during our 2015 transaction testing. We noted that seven out of the eight had already updated their action plans by February 2016, including measures to address some of the issues we found. During our examination of Member States’ systems, however, we found that the four paying agencies where we had detected weaknesses in the administrative checks on public procurement procedures had not included any action on this issue in their action plans. While we consider that the action plans are generally a sound tool to address the causes of errors, there is still scope for further improvement in the area of public procurement. When weaknesses are identified in the framework of conformity audits and audits by the Court, the Commission always requests that Member States take corrective actions, which may take the form of an action plan. Following the Court’s recommendations in special report No 23/2014, the Commission has recommended all Member States to pay particular attention to the issue of public procurement and, where necessary, take remedial actions. Reservations were made in the 2015 annual activity report to cover public procurement issues, as recognised by the Court in point 7.60. Member States' systems related to cross-compliance We examined the system for ensuring that cross-compliance requirements are met in the six paying agencies visited (see paragraph 7.32). In particular, we analysed the reliability of control statistics and the procedures put in place to ensure that Member States appropriately check cross-compliance requirements for protecting waters against nitrates (the Nitrates Directive) and for cattle notifications. We also reviewed 63 control reports. The Commission notes that cross-compliance does not relate to assessment of the legality and regularity of underlying transactions. We found deficiencies, mainly relating to the reliability of control statistics and sampling, in five out of the six Member States visited. We found problems (e.g. incorrect calculation of sanctions or lack of checks on certain requirements) in 13 of the 63 control reports reviewed. Moreover, in Lithuania and the UK (Northern Ireland) in four of the 24 control reports sampled the sanctions were not correctly included in the control statistics. We give another example of the deficiencies we identified in Figure 7.10. The Commission shares the Court’s observation and pays particular attention to these requirements during its audits. In many of them, the Commission has observed weaknesses in the scope and quality of cross-compliance controls. When a systemic non-compliance is established, the Commission always requests corrective actions, including actions plans whenever necessary and ensures the protection of the EU budget via the conformity clearance procedure. Example of insufficient cross-compliance checks In Lithuania we found that national authorities did not sufficiently check the cross-compliance requirements for protecting waters against nitrates, which are applicable to all farmers. The Member State used several thresholds, under which farmers were not subject to cross-compliance checks for requirements resulting from the Nitrates Directive. For example, farmers with a utilised agricultural area of 10 hectares or less do not need to declare the level of fertiliser used. Such farmers account for more than 70 % of all farmers and use 14 % of the agricultural area. As another example, livestock farmers who keep fewer than 10 livestock units in one place do not need to have a manure storage of the relevant capacity. Such farmers constitute 87 % of all livestock farmers and keep 28 % of all livestock units in Lithuania. The Nitrates Directive aims at protecting the environment and citizens’ health. While the use of thresholds may reduce red tape for smaller farmers, excluding farmers from cross-compliance checks is not only irregular, but may also adversely affect the achievement of the objectives of the Nitrates Directive. Example of insufficient cross-compliance checks The Commission takes note of the Court’s finding. Although the Commission’s 2014 audit in Lithuania only covered animal-related SMRs, similar deficiencies were found in terms of sampling. Those deficiencies are being followed up in a conformity clearance procedure. DG MARE and Member States systems related to the regularity of fisheries transactions Managing authorities make payments to the beneficiaries, and declare the expenditure to the Commission, which then reimburses the Member States. The Commission uses the annual control reports and the annual opinion of the audit authorities to assess whether the Member States have set up management and control systems that comply with the legal requirements. The Commission also carries out its own audits to verify that the systems function effectively. Our examination of systems in this area focused on audits carried out by the Commission (DG MARE). We carried out a desk review of a sample of five of the Commission's conformity audits. The Commission reported significant weaknesses in the systems of two (Italy and Romania) out of the five Member States concerned. We reviewed the Commission’s audit work and its compliance with the international auditing standards. The five audits which we reviewed were not fully compliant with the standards, and we identified weaknesses in audit supervision and audit documentation. The Commission recognised that improvements were required in the areas of audit supervision and documentation, and strengthened its procedures accordingly. The files audited predate these changes. We also re-performed one Commission audit in Romania. We identified deficiencies in the managing authority’s administrative checks, in particular relating to procurement procedures (see Figure 7.11). Procurement deficiencies in Romania We examined a sample of seven procurement procedures, five of which were affected by errors (two public and three private). The three private cases were not detected by the Commission. We also identified deficiencies relating to the project selection procedure in two out of six cases reviewed. Procurement deficiencies in Romania The standard checklists used by the Commission focus on compliance with public, rather than private, procurement rules. The seven procedures referred to concern both private procurement cases, on the basis of instructions drawn up by the Romanian authorities (five cases) and public procurement cases (two cases relating to technical assistance) where the Commission was already aware of the weaknesses. Furthermore, the Commission had previously identified that there were weaknesses in project selection. On the basis of the 24 audits which it carried out in 2014 and 2015 and which were finalised at the time of our audit, DG MARE proposed two financial corrections. Our review of a sample of five of these audits revealed that in two cases (Italy and Romania) the financial corrections did not address all the deficiencies affecting the managing authorities’ internal control systems. The Commission proposes financial corrections where it identifies significant deficiencies. It considers that appropriate levels of correction were applied for both Italy and Romania based on its own analysis. However, for Romania, it will assess the need for additional corrections based on the Court’s findings. DG AGRI’s annual activity report In its 2015 AAR DG AGRI assesses the expenditure at risk for the main spending areas. The starting point of this assessment is the levels of error reported by the Member States (for each paying agency) in their control statistics. Subsequently, the Commission adjusts those levels of error using mainly the Commission's and the Court's audits carried out in the last three years as well as the certification bodies’ opinions on legality and regularity for the 2015 financial year. The resulting adjusted error rates (AERs) are then aggregated and used to calculate the total expenditure at risk, i.e. the level of error that the Commission estimates to affect the EU budget. We consider that this is generally a valid approach which can provide sufficient basis for reservations at the level of individual paying agencies. However, we highlight the following issues. For the first time in 2015, certification bodies were required to ascertain, on the basis of a representative sample, the legality and regularity of the expenditure for which reimbursement has been requested by the paying agencies from the Commission. The certification bodies thus represent a key element in the Commission assurance model for the 2014-2020 period. In 2015, DG AGRI could use the opinions of the certification bodies on the legality and regularity only to a very limited extent due to significant weaknesses in methodology and implementation (such as delays in appointing certification bodies in Member States, inadequate audit strategies, too small samples being drawn for substantive testing and insufficient technical skills and legal expertise to detect deficiencies in paying agencies’ administrative and on-the-spot checks). The Commission considers that having an opinion on legality and regularity by an independent, professional audit body at national level will, over time, result in an improvement in the management and control systems of the Member States and a greater level of assurance on the error rates reported by the Member States. As financial year 2015 was the first year of implementation of this new approach, the full benefits of the new work of the certification bodies could not yet be reaped. Nevertheless, a number of encouraging messages and lessons for the future can be taken from the 2015 exercise. The Commission made adjustments to determine the AER on the part of expenditure for which its own or our audit results were available. However, DG AGRI did not disclose in its AAR which part of the expenditure was not adjusted because no such audit results were available. In such cases, the Commission places reliance on Member States control statistics which may result in an underestimated AER. The Commission examines the situation for each paying agency not only based on its own and the Court’s audit findings over the three previous years, but also on the basis of the opinions received for the financial year concerned from the certification body for each paying agency. There are therefore audit results available in all cases. For the first time in 2015, certification bodies were requested to provide an opinion on the legality and regularity of underlying transactions, which provided even more details and information relevant for deciding on adjusting the reported error rates. THE COURT’S OBSERVATIONS Figure 7.12 gives the key data used to assess DG AGRI’s 2015 AAR. Source: European Court of Auditors based on DG AGRI annual activity reports. Main spending area AAR (financial year) 2015 expenditure (million euro) Average level of error reported by the Member States Aggregated adjusted error rate Corrective capacity Average financial corrections over the last three years Average recoveries over the last three years Total (million euro) (million euro) EAGF 2014 44137,85 0,55 % 2,61 % 536,44 1,22 % 117,83 0,27 % 1,49 % 2015 44834,52 0,68 % 1,47 % 665,36 1,48 % 107,41 0,24 % 1,72 % EAFRD 2014 11186,00 1,52 % 5,09 % 113,58 1,02 % 95,64 0,85 % 1,87 % 2015 9843,14 1,78 % 4,99 % 177,99 1,81 % 114,95 1,17 % 2,98 % THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES For EAGF, in spite of an increase in the average level of error reported by the Member States from 0,55 % in 2014 to 0,68 % in 2015, the Commission concludes that the adjusted error rate has fallen from 2,61 % in 2014 to 1,47 % in 2015. The Commission explained that this reduction of the adjusted error rate is largely caused by:

improvements in the Member States’ control systems as a result of remedial actions implemented as from the 2014 claim year;

more up-to-date information available to the Commission allowing a reduction of previously calculated adjustments.

The Commission considers that an increase in the average level of the reported errors proves that the Member States’ management and control system work better at detecting and reporting errors. It follows that the reported error rate is considered more reliable and therefore fewer adjustments are needed. The factors mentioned by the Court have positively impacted the evolution of the adjusted error rate. For EAGF, the AAR reports that the aggregated adjusted error rate has fallen from 2,61 % in 2014 to 1,47 % in 2015. Our audit results show that the estimated level of error remained stable, when excluding quantification of cross-compliance errors (see Annex 7.1). For EAFRD, the Commission almost tripled the average level of error reported by the Member States of 1,78 % to calculate the adjusted error rate of 4,99 %. The Commission has adjusted the error rates reported by 56 paying agencies out of 72 declaring EAFRD expenditure. We consider the adjusted error rate of 4,99 % to be broadly in line with our estimated level of error. The Commission has in recent years increased the number of its conformity audits of EAFRD expenditure and also launched horizontal audits targeting measures/paying agencies with similar risks. An example is an audit of public procurement which contributed to overall reservations being made for Germany and Spain. The Commission defines corrective capacity as a mechanism for correcting the errors made and protecting the EU budget. The annual increase in the calculated corrective capacity for EAGF (to 1,72 %) and for EAFRD (to 2,98 %) is the result of recent efforts by the Commission’s to reduce the number of old open conformity audits (see paragraph 7.34). This affected the overall amount of the financial corrections adopted, especially in 2015. The corrective capacity is an estimate of the amounts of future corrections calculated as a historical average of the net financial corrections as explained in the Directorate-General for Agriculture and Rural Development’s 2015 annual activity report. The Commission corrective capacity is calculated on the basis of a three year average of the financial corrections received and has no direct link with the 2015 expenditure which will be corrected in the years to come. This explains this year’s anomaly that the Commission reports its corrective capacity for EAGF expenditure to be higher than the adjusted error rate. For EAFRD, there is no assurance that the remaining financial risk will be below materiality by the end of the programming period. The Commission confirms this in its 2015 AAR. The corrective capacity and the adjusted error rate are two estimates which provide together an indicator of the assurance on the effectiveness of the multiannual control system in protecting the EU budget. In its 2015 annual activity report, the Director-General for Agriculture and Rural Development gives assurance for the 2015 CAP expenditure as a whole, taking into account both the adjusted error rate and the corrective capacity. DG MARE’s annual activity report We examined DG MARE’s 2015 annual activity report, focusing on whether all substantial information was disclosed, whether the information provided was in line with our own findings, and whether the estimations used by the Commission to calculate the error rates were reasonable. DG MARE sent out three pre-closure letters for three operational programs proposing to recover an amount of 53,6 million euro. This material amount is not disclosed in the AAR. The AAR includes reservations for five Member States, including Romania, which is in line with our audit findings. Our audit did not call into question the calculation of the error rates by DG MARE. This amount was not disclosed as the Commission does not consider the amount to be at risk. A recovery order has been issued for one of the three programmes, for the amount anticipated. For the remaining two programmes, when the annual activity report was drawn up the Member State and the Directorate General for Maritime Affaires and Fisheries already largely agreed on the amounts to be recovered by the Commission. Any adjustments to this amount (and the associated financial risk) will therefore be minimal. Conclusion and recommendations The conclusion for 2015 Overall evidence indicates that spending on Natural resources and on both areas subject to a specific assessment (see paragraphs 7.13 and 7.21) is affected by a material level of error. For this MFF heading area, testing of transactions indicates that the estimated level of error present in the population is 2,9 % (see Annex 7.1). The Commission takes note of the most likely error estimated by the Court which is broadly the same as last year excluding cross-compliance. Recommendations Annex 7.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, the Court presented 20 recommendations. The Commission and the Member States fully implemented four recommendations, while five were implemented in most respects and 10 in some respects. The Commission takes note of the latest assessment made by the Court of the progress in addressing recommendations made in previous annual reports. The Commission will continue working with the Member States to further improve the way they follow up the Court’s recommendations. Following this review and the findings and conclusions for 2015, we recommend that the Commission, for EAGF:

Recommendation 1: continue its efforts to follow up on cases where national legislation is not compliant with EU legislation, including all legal means at its disposal, in particular suspension of payments (see paragraph 7.20 and Figure 7.5);

The Commission accepts the recommendation. The Commission considers that it is being implemented by means of its ongoing audit work and the provision of guidance when required and takes the necessary steps in the context of shared management to assist the Member State audit bodies to discharge their tasks effectively. In addition, shortcomings in Member States management and control systems are addressed through targeted and comprehensive action plans where necessary. When the Commission detects such problems during the course of its monitoring or audits, they request the Member State to take remedial actions. Where the problem is particularly acute, the Member State is required to implement a remedial action plan which is closely followed by the services. So far such plans have been found to be very effective. When the implementation of the action plan is not deemed to be effective, payments can be reduced or suspended in order to protect the EU budget.

Recommendation 2: annually monitor the results of the LPIS quality assessments performed by Member States, and check that all Member States with negative assessments actually take the necessary remedial action (see Figure 7.9);

The Commission accepts this recommendation. The results of the LPIS quality assessment (QA) are monitored on an annual basis. This yearly exercise is a quality control process whose ultimate objective is to improve the quality and reliability of the LPIS, by analysing detected technical weaknesses and to taking remedial actions to address them. for rural development:

Recommendation 3: ensure that all Member States’ action plans addressing errors in rural development include effective actions on public procurement (see paragraph 7.45);

The Commission accepts this recommendation. The Commission is fully engaged in providing targeted and effective guidance on how to prevent irregularities in public procurement procedures: Public Procurement — a study on administrative capacity of the EU (2016), including country-specific recommendations, and Guidance for practitioners on the avoidance of the most common errors in public procurement of projects funded by the European Structural and Investment Funds (2015). In December 2015 the Commission has underlined that all Member States need to pay particular attention to public procurement rules, and take corrective action where relevant. The Commission is working intensively with the Member States in order to identify and remedy the causes of errors in the implementation of the policy. This work will continue with the aim to further address identified shortcomings, including requesting action plans. The Commission insists on the completeness and effectiveness of these action plans. Member States are requested to take corrective actions when weaknesses are identified in the context of conformity audits, audit by the ECA or findings of the certification body or paying agency. Finally, net financial corrections are applied where necessary. for EAGF and rural development:

Recommendation 4: monitor and actively support certification bodies in improving their work and methodology on the legality and regularity of expenditure so that they provide reliable information for the Commission’s estimation of the adjusted error rate (see paragraph 7.54);

The Commission accepts this recommendation. The Commission will continue to support the Member States to improve and consolidate the work done by the certification bodies, to ensure that, in the coming years, increased assurance can be drawn from their work.

Recommendation 5: update DG AGRI’s audit manual by including detailed audit procedures and documentation requirements for the verification of the data supplied by Member States and used for calculating financial corrections (see paragraph 7.36);

The Commission accepts this recommendation. The Commission will when updating its audit manual address this issue building on the experience gained. for fisheries:

Recommendation 6: improve the compliance of DG MARE’s conformity audits with international auditing standards (see paragraph 7.50).

The Commission accepts this recommendation. It should be noted that the procedural improvements that took place after the period covered by the Court’s audit have already addressed many of the issues highlighted here (in particular as regards audit evidence and audit trail). Furthermore, a quality control procedure is currently being activated. PART 2 — RURAL DEVELOPMENT PERFORMANCE-RELATED ISSUES Assessment of projects’ performance Last year, we reported our work on performance-related issues for selected rural development transactions. This year, we continued this work. Of the 159 rural development transactions tested for regularity, we assessed performance related issues for 59 investment projects in tangible assets. Of the 59 investment projects 34 were completed at the time of our audit visit. For these completed investment projects we checked whether the investment had been carried out as planned. This criterion was fulfilled for most projects (95 %). For the 59 investment projects we found the following main performance issues:

insufficient evidence that costs are reasonable;

deficiencies in targeting measures and selecting projects, including weak links to Europe 2020 objectives.

Insufficient evidence that costs are reasonable For 33 of the 59 investment projects (56 %), we found sufficient evidence that project costs were reasonable. However, for 26 investment projects (44 %), this was not the case. Last year the respective figures were 69 % and 31 %. For both years, the projects examined were implemented in the 2007–2013 programming period and under the same rules. Figure 7.13 gives two examples of investment projects for the same rural development measure, one where we found that costs were reasonable, and one where this was not the case. In 2014 we published a special report containing a detailed examination of this topic. The assessment of reasonableness of costs is the responsibility of the Member States (see article 48 of Regulation (EU) No 809/2014). The Commission protects the EU budget when weaknesses are identified through conformity procedures and issues guidance on how to improve the systems in place (see Guidance on rural development controls and sanctions, including a checklist for the assessment of reasonableness of costs). Furthermore, training and sharing of good practices are organised by the European Network of Rural Development. Finally, the Commission is actively promoting the use of simplified cost options in the Member States. Examples of checking reasonableness of costs for the modernisation of agricultural holdings In Greece, a beneficiary bought a tractor partially financed by rural development funds. The beneficiary had to use a predetermined formula to calculate the appropriate horsepower of the tractor, taking into consideration the cultivated area of the land. The beneficiary provided an offer which the paying agency verified against a reference price which was taken from a specialist magazine about farm machinery. The paying agency concluded that the price for the specific model should be 10 % lower than the price proposed. As a result, the beneficiary bought a cheaper tractor. In Slovakia, we audited a project for the construction of a storage facility. The paying agency did not check if the project costs were reasonable. However, we obtained conclusive evidence that the price of one of the main construction materials (concrete) was six times higher than the normal market price. Overall, the cost of the project should have been more than 50 % lower. Examples of checking reasonableness of costs for the modernisation of agricultural holdings For the Slovakian case the Commission notes that the project was implemented. Out of the 59 investment projects, 50 (85 %) were in line with the needs for both quantity and quality. For example, we visited three projects that relate to the construction or the extension of commune centres in small towns or villages under the measures village renewal and development and implementing local development strategies. The investments were fully utilized at the time: all the rooms were in use and activities for the local community were scheduled for weeks ahead. However, for nine projects (15 %) the output of the project was larger than actually needed. One example of such a case is the construction of a warehouse which was several times bigger than needed and therefore remained for a large part unused. Another example of a project which included non-essential items is given in Figure 7.14. Example of a project which included non-essential items In Italy (Campania), a municipality renewed a 1-kilometre footpath in a mountainous area for a cost of 441000 euro. During the procurement procedure, the municipality asked interested companies for proposals to make the investment more functional and environmentally friendly. The proposal by the company selected included additional items costing 80000 euro, over and above the cost of the work on the footpath. These additional items included a mountain bike worth 4000 euro, a panoramic spyglass costing 3500 euro, and a 10000 euro donation to a local church. Instead of restricting the contract to the needs of renovating the footpath, additional items were added which led to higher costs. Example of a project which included non-essential items The project in question, supported under the forestry measure 227 Non-productive investments, is under investigation by the Region. The Regional Managing Authority has not paid the final balance of the project yet and intends to make a reduction as a result of the control carried out. Deficiencies in targeting the measures and selecting projects We tested whether support was targeted at clearly defined objectives reflecting structural and territorial needs and structural disadvantages. This criterion was not fulfilled for 22 % of the investment projects examined. Last year we reported a very similar percentage (21 %). Examples of good and bad practice are given in Figure 7.15. We have reported on the lack of adequate targeting for rural development investments in our special reports. For the 2014-2020 period, selection criteria have to be established for all measures but the area- and animal-related ones. The principles of selection criteria, in line with strategy and objectives of the policy, must be set in the programme. Applications will then have to be assessed and ranked against specific selection criteria on which the Monitoring Committee should be consulted. Minimum thresholds should be established for the application of selection criteria to avoid financing poor projects. Examples of targeting and selection In Italy (Campania), the paying agency opened a call for proposals with a budget of 25 million euro under the measure modernisation of agricultural holdings. The paying agency applied different selection criteria as regards the beneficiary (e.g. criteria on young farmers, gender balance), the holding (e.g. criteria on holdings situated in less favoured areas or producing quality products) and the project (e.g. criteria on synergies with other measures, innovation etc.). The Italian authorities also set a minimum threshold in advance in order for a project to be selected. In this case, the Italian authorities prioritised the principles of sound financial management above spending. In Hungary, the objectives of the measure modernisation of agricultural holdings include three broad areas (upgrading, diversification and compliance with standards). The audited sub-measure was targeted towards animal holdings with a certain minimum size. There was no other prioritization or geographical targeting of the aid. The rural development programme split the available budget into amounts that become available periodically to applicants. Although the Hungarian authorities defined selection criteria, they did not use them. The call for proposals for the audited project opened on 16 December 2013, for a budget of 10 million euro. The Hungarian authorities funded projects on a first come, first served basis. The last project accepted was submitted on 19 December 2013, only three days after the opening of the call. In this case, the selection process took no account of the projects’ potential impact. The approach observed in Hungary allows the paying agency to fund projects with lower value-added, for the simple reason that they are submitted earlier than other projects with higher added value. Examples of targeting and selection The added-value of projects is taken into account as part of the eligibility checks. At the very end of the programming period 2007-2013, in December 2013, Hungarian authorities did not use the same selection system as during the previous years. The objective was to support investments that can be still finished before the end of 2015. For the 59 investment projects examined, we assessed whether Europe 2020 targets such as innovation or renewable energy were translated into operational elements for targeting and selection procedures. We found this was the case in only 51 % of the projects reviewed, despite the fact that almost all of them (57 out of 59) had been approved after the adoption of Europe 2020 targets. Examples of good and bad practice are given in Figure 7.16. Projects selected by national authorities during the 2007-2013 programming period should not be evaluated against Europe 2020 objectives as those had not been defined when the legal framework for the programming period was adopted (in 2005). In that light, the Commission considers it a positive result that for 51 % of the projects reviewed Europe 2020 targets were reflected. The Commission considers that rural development priorities as set out for 2014-2020 are a translation of thematic objectives of Europe 2020 into the specific context of the rural development policy. EAFRD legal framework establishes a clear link: the contribution to Europe 2020 shall be pursued through six priorities subdivided into focus areas. This requirement is fully reflected in the structure of the EAFRD programmes. The progress towards targets set for the priorities/focus areas is monitored in the Annual Implementation Reports. In order to ensure consistency with the thematic objectives, correspondence tables have been annexed to the guidance documents. Examples of links to EU priorities In Italy (Campania), we audited a project with a value of almost 100000 euro under the measure non-productive investments. The project consisted in rebuilding dry stone walls with stones recovered from a previously existing wall of a similar size, aimed at reconstructing elements of the rural landscape. For this case, we did not identify a link to any of the objectives of the EU 2020 Strategy. In this region we also audited a project which was selected on the basis of the procedure described in Figure 7.15. The project included investments in a new irrigation system, an area for rain water collection and photovoltaic panels. The project contributes to adaptation to climate change and energy efficiency, which is one of the five headline targets of the EU 2020 Strategy. Examples of links to EU priorities Rebuilding dry stone walls is an action covered, under the 2007-2013 rural development programming period, by measure 216, with the objectives of land protection and conservation of biodiversity. The investment is relevant in relation to those objectives and in line with the EU biodiversity strategy to 2020 (COM(2011) 244 final). Biodiversity, renewable energy and water management are closely linked to each other and can all contribute to the climate change challenges. Conclusion Our examination of performance aspects for rural development investment projects showed that most investments were carried out as planned. However, we found deficiencies in targeting and a lack of evidence that costs were reasonable. In the projects examined, Member States did not sufficiently take EU priorities into account. For these issues we did not notice an improvement compared to last year. The Commission has identified similar weaknesses in the Member States’ assessment of reasonableness of costs in the framework of audit conformity procedures, applying where relevant financial corrections to protect EU budget. Furthermore, the Commission has published guidance on controls and penalties under rural development, including the checklist to assess reasonableness of costs developed by the Court of Auditors in the special report No 22/2014. Finally, the Commission is actively advocating the use of simplified cost options under investment measures, which would contribute to the avoidance of such type of irregularities. As regards targeting, the Commission notes that this is ensured through the intervention logic of the policy and the programmes: EAFRD-supported operations (projects) must comply with objectives of the measure programmed in the RDP, applying adequate selection criteria. The measures contribute to the rural development priorities, which are consistent and reflect EU2020 strategy for smart, sustainable and inclusive growth (see Article 5 of Regulation (EU) No 1305/2013). The common monitoring and evaluation system facilitates tracking of results and evaluates impacts in overall strategies (CAP and EU2020).

ANNEX

Anx. 71. 2015 2014 2014 without cross-compliance SIZE AND STRUCTURE OF THE SAMPLE Agriculture: Market and direct support 180 183 Rural development, environment, climate action and fisheries 179 176 Total transactions Natural Resources 359 359 ESTIMATED IMPACT OF QUANTIFIABLE ERRORS Estimated level of error: Market and direct support 2,2 % 2,9 % 2,2 % Estimated level of error: Rural development, environment, climate action and fisheries 5,3 % 6,2 % 6,0 % Estimated level of error: Natural Resources 2,9 % 3,6 % 3,0 % Upper Error Limit (UEL) 4,5 % Lower Error Limit (LEL) 1,3 %

ANNEX 7.1 RESULTS OF TRANSACTION TESTING FOR NATURAL RESOURCES

2015 2014 2014 without cross-compliance SIZE AND STRUCTURE OF THE SAMPLE Agriculture: Market and direct support 180 183 Rural development, environment, climate action and fisheries 179 176 Total transactions Natural Resources 359 359 ESTIMATED IMPACT OF QUANTIFIABLE ERRORS Estimated level of error: Market and direct support 2,2 % 2,9 % 2,2 % Estimated level of error: Rural development, environment, climate action and fisheries 5,3 % 6,2 % 6,0 % Estimated level of error: Natural Resources 2,9 % 3,6 % 3,0 % Upper Error Limit (UEL) 4,5 % Lower Error Limit (LEL) 1,3 %

(1) the eligibility and size of agricultural parcels, in particular of permanent pasture, are correctly assessed and recorded by the Member States on the basis of the applicable EU criteria by way of comprehensive analysis of the most recent orthoimages (see paragraphs 3.9, 3.10 and 3.20); (2) immediate remedial action is taken by the Member States where the IACS is found to be affected by systemic errors especially as regards incorrect administrative treatment of aid applications (see paragraph 3.15 and Box 3.5); (3) the debtors ledgers of the Member States contain full and reliable information on the amounts and nature of debts and that effective recovery/enforcement procedures are applied without any undue delay (see paragraphs 3.24 and 3.25); Anx_72. 346. Year Court Recommendation Court's analysis of the progress made Commission reply Fully implemented Being implemented Not implemented No longer applicable Insufficient evidence In most respects In some respects 2013 Following this review and the findings and conclusions for 2013, the Court recommends that: Recommendation 1: the Commission and Member States take the necessary measures to ensure that the IACS is used to its full potential. This comprises particular efforts in ensuring that: X The Commission is monitoring the situation: where weaknesses persist, action plans are being implemented, and conformity clearance procedures are ongoing. In its 2014 annual report, the Court did not address any recommendation to the Commission regarding IACS, and the same is true for the 2015 annual report. 2013 Recommendation 2: the Commission ensures that the reinforcement of assurance procedure is effectively applied in order to enhance the quality and comparability of the work performed by the audit bodies and that remedial action is taken in respect of the unjustified reduction by the Italian authorities of the on-the-spot inspection rate (see paragraphs 3.30 to 3.35); X The Commission considers the recommendation as implemented in most respects. The reinforcement of assurance procedure as such does not exist anymore. The Commission has introduced guidelines for the certification bodies in order to provide a reliable basis for the assessment of legality and regularity of expenditure. 2015 was the first year in which the certification bodies were required to produce an opinion on the legality and regularity of transactions. The Commission will build on the experience of this first year. Regarding specifically the weaknesses identified in Italy, the ongoing conformity procedure will ensure that any damage to the EU financial interest will be corrected. Recommendation 3: the Commission actively monitors the application of remedial actions with regard to the deficiencies in the control system applicable to EU aid for producer groups in Poland (see paragraphs 3.36 to 3.38). X The Commission considers the recommendation as fully implemented. A remedial action plan is ongoing and is being closely monitored by the Commission. Delays in the implementation have triggered a suspension of the payments. An ongoing conformity procedure will ensure that the EU financial interest is protected. 2013 Following this review and the findings and conclusions for 2013, the Court recommends in the area of rural development that: Recommendation 1: the Member States carry out their existing administrative checks better, by using all relevant information available to the paying agencies, as this has the potential to detect and correct the majority of errors (see paragraphs 4.8 and 4.20). In particular for investment measures, administrative checks should use all available information to confirm the eligibility of the expenditure, project and beneficiary (including all ultimate shareholders) and compliance of public procurement procedures with the applicable EU and/or national rules; X The Commission discusses measures to improve the quality of administration and control systems with the Member States in meetings with the learning network, in DG AGRI’s simplification working groups, the Farm Advisory System (FAS), annual screening of LPIS quality and also in the conferences with the Directors of the PAs. The new legal framework for CAP 2014-2020 contains provisions which reduce risk of error. For rural development, various measures have been merged and streamlined, reducing the total number of measures. In this context, provisions have been clarified, eligibility rules have been modified in order to be easier to implement and contract conditions have been made more flexible. Simplified costs options, i.e. standard scales of cost units, lump sums and flat-rate financing can be applied so that the processes of claiming, administering and auditing reimbursement for payments made will be easier for everyone. Public procurement is one of the central elements in the action plans for rural development. Nevertheless the non-compliance of procurement does not necessarily mean that 100 % of the expenses involved are ineligible. The project as such may well achieve its goal and provide real added value. Moreover, it is noted that new guidelines for determining financial corrections to be made to expenditure financed by the Union under shared management, for non-compliance with the rules on public procurement, have been recently adopted (Commission Decision C(2013) 9527). 2013 Recommendation 2: the Commission ensures that all cases where the Court detected errors are followed up appropriately (as identified in paragraphs 4.14, 4.20 and 4.21, as well as in paragraph 4.16 for environment). X The Commission takes note that ECA’s analysis is on-going. Following the ECA remarks, in the latest updates of action plans the Commission and the ECA audit findings are more systematically followed-up. It should also be noted that all material DAS cases reported by the Court are followed up by the Commission, including through conformity procedure, if appropriate. The number of audits carried out by the Commission on EAFRD expenditure has also increased very substantially. National action plans were requested from the Member States and their implementation is being closely monitored by DG AGRI services. In the latest updates of the action plans, the Commission made sure that its own and ECA’s audit findings were more systematically followed up. Actions are time-framed and monitored through appropriate indicators. Geographical desks follow up the root causes of errors and mitigating actions in monitoring committees, annual review meetings with the managing authorities, bilateral meetings and programme amendments when relevant. Audit findings are communicated to geographical desks on a regular basis in order to allow them to discuss with Member States in view of improving the implementation of programmes. The Commission is organising through the European Network for Rural Development (ENRD) specific workshops on relevant issues aiming to reduce the error rate and improve the implementation of the policy. 2013 Recommendation 3: the Member States ensure that action plans to address the high error rate in rural development are complete, by including all regions and addressing all measures, particularly investment measures, and take the Commission's and Court's audit findings into account (see paragraph 4.30). X The Commission has already put in place a response to the high error rate. The Commission systematically recommends as part of the clearance procedure that Member States improve the control systems and ensure a more vigorous implementation thereof. Moreover, the Commission has asked systematically the paying agencies concerned by reservations to put in place action plans to remedy the shortcomings behind the reservations. When shortcomings are identified in the Member States’ management and control systems, action is taken at two levels: first through the opening of conformity clearance procedures to cover the financial risk for the EU budget; secondly, whenever necessary and appropriate, by asking the Member States to set up action plans to remedy the deficiencies. National action plans were requested from all Member States in January 2013 and their implementation is closely monitored by DG AGRI services (cf. the Commission staff working document on the assessment of root causes of errors in the implementation of rural development policy and corrective actions (SWD(2013) 244)). The Court has recognised the value of these measures and considers that they are a step in the right direction in addressing the causes of errors (see special report No 23/2014 on errors on rural development spending). In the latest updates of the action plans, the Commission made sure that its own and ECA’s audit findings were more systematically followed-up. In the recommendations made in special report No 23/2014, the Court stated that following preventive and corrective actions should be taken by the Member States where relevant. The Commission therefore does not consider that action plans have to be made for all regions and all measures but only where deficiencies have been identified. 2013 and for the CAP as a whole that: Recommendation 4: the Commission documents how it calculates the expenditure covered by its conformity audits (see paragraph 4.24). X The Commission will continue to report on the audit coverage in its AAR. Recommendation 5: the Commission takes steps to further reduce the backlog of open audit files, so as to enable all audits carried out prior to 2012 to be closed by the end of 2015 (see paragraph 4.27). X At the end of June 2016, only 20 files remained in the backlog. Recommendation 6: the Commission further develops its approach to calculating the RER by ensuring that it takes into account all expenditure and paying agencies (see paragraph 4.29). X The Commission considers the recommendation as fully implemented. The methodology for estimating the adjusted error rate is detailed in the Annex 4 to DG AGRI’s AAR. Moreover, following an IAS recommendation, detailed procedures on how the estimation of the remaining amount at risk has been drawn up (c.f. Ares(2014) 2509210 — 29/07/2014). These procedural steps were updated in 2015 and complemented by additional documentation on the conceptual model underlying the DG AGRI’s estimation of the error rate at the level of each paying agency. Furthermore, the Court recommends that: Recommendation 7: financial corrections to Member States concerning the European Fisheries Fund are supported by evidence of their validation (see paragraph 4.34). X 2012 Following this review and the findings and conclusions for 2012, the Court recommends that the Commission and Member States increase and speed up their efforts to ensure that: Recommendation 1: the eligibility of land, and in particular permanent pasture is properly recorded in the LPIS, especially in cases where areas are fully or partly covered with rocks, shrubs or dense trees or bushes or where land has been abandoned for several years. X The Commission is monitoring the situation: where weaknesses persist, action plans are being implemented, and conformity clearance procedures are ongoing. In its 2014 annual report, the Court did not address any recommendation to the Commission regarding IACS, and the same is true for the 2015 annual report. Recommendation 2: immediate remedial action is taken where administrative and control systems and/or IACS databases are found to be deficient or out of date. X When the Commission services detect such problems during the course of their audits, they request the Member State to take remedial actions. Where the problem is particularly acute, the Member State is required to implement a remedial action plan which is closely followed by the services. So far such plans have been found to be very effective. The Commission also ensures that the financial risk to the EU budget arising from such deficiencies is covered via financial corrections imposed via the conformity clearance procedure. Recommendation 3: payments are based on inspection results and that on-the-spot inspections are of the quality necessary to determine the eligible area in a reliable manner. X The Commission considers the recommendation as implemented in most respects .The Commission shares the view of the Court in this regard and it will continue to focus its audits on the risk of ineligible land being not detected prior to the payments. Financial corrections will be applied where necessary to protect the EU financial interests. 2012 Recommendation 4: the design and quality of the work performed by the directors of paying agencies and the certification bodies in support of their respective declarations and statements provide a reliable basis for the assessment of the legality and regularity of underlying transactions. X The Commission continues to review the work of the certification bodies via the analysis of the documents related to the financial clearance of accounts as well as audit missions to certification bodies. In addition, the guidelines necessary for the certification bodies to complete their work on the certification of the accounts for financial year 2015, which for the first time will include an assessment of the legality and regularity of the underlying transactions, were finalised in 2014 and the final version officially presented to Member States in January 2015. This included the updated guideline dealing with the revised management declaration. Since 2012 four expert group meetings took place where these guidelines were presented and discussed. In order to gain further assurance on the quality of the work of the CBs 15 missions took place in 2015 that will assess the CB’s work on legality and regularity, in particular focusing on the quality of their on-the-spot controls. Additional expert group meetings took place in 2015 and 2016 where experiences of the work on the legality and regularity of transactions were shared. Following this review and the findings and conclusions for 2012, the Court recommends in the area of rural development that: Recommendation 1: the Member States carry out their existing administrative checks better, by using all relevant information available to the paying agencies, as this has the potential to detect and correct the majority of errors. X See Commission’s reply to recommendation 1 for rural development from 2013. Recommendation 2: the Commission ensures that all cases where the Court detected errors are followed up appropriately. X 2012 Recommendation 3: the Commission, in DG AGRI's annual activity report, applies a similar approach for EAFRD as for decoupled area aid, where the Commission takes account of the results of its own conformity audits in assessing the error rate for each paying agency. X and for the CAP as a whole that: Recommendation 4: the Commission ensures adequate coverage in its conformity audits. X The Commission seeks to increase the audit coverage within the limits of the resources available. In the framework of its multiannual strategy, DG AGRI is applying from July 2014 a rolling three-year audit programme based on a central risk analysis and designed in a way to maximise the audit coverage. Recommendation 5: the Commission addresses the weaknesses identified in its conformity audits and the persistent problem of long delays in the conformity procedure as a whole. X The Commission will continue to improve its audit methodology and procedures. Moreover, it will continue to make efforts, notably in the framework of the implementation of the CAP reform, to improve and speed up the process bearing in mind the need to maintain quality standards and the Member State’s right of reply. Commission Implementing Regulation (EU) No 908/2014, applicable from January 2015, introduced compulsory deadlines for the different steps of the audit procedure. Reinforced monitoring tools have also been introduced. The backlog of files has almost been cleared by the end of June 2016. Recommendation 6: the Commission further improves its method of determining financial corrections so as to take better account of the nature and gravity of the infringements detected. X

ANNEX 7.2 3.46.4.37.4.38.3.37.4.44. FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR NATURAL RESOURCES

Year Court Recommendation Court's analysis of the progress made Commission reply Fully implemented Being implemented Not implemented No longer applicable Insufficient evidence In most respects In some respects 2013 Following this review and the findings and conclusions for 2013, the Court recommends that: Recommendation 1: the Commission and Member States take the necessary measures to ensure that the IACS is used to its full potential. This comprises particular efforts in ensuring that: X The Commission is monitoring the situation: where weaknesses persist, action plans are being implemented, and conformity clearance procedures are ongoing. In its 2014 annual report, the Court did not address any recommendation to the Commission regarding IACS, and the same is true for the 2015 annual report. 2013 Recommendation 2: the Commission ensures that the reinforcement of assurance procedure is effectively applied in order to enhance the quality and comparability of the work performed by the audit bodies and that remedial action is taken in respect of the unjustified reduction by the Italian authorities of the on-the-spot inspection rate (see paragraphs 3.30 to 3.35); X The Commission considers the recommendation as implemented in most respects. The reinforcement of assurance procedure as such does not exist anymore. The Commission has introduced guidelines for the certification bodies in order to provide a reliable basis for the assessment of legality and regularity of expenditure. 2015 was the first year in which the certification bodies were required to produce an opinion on the legality and regularity of transactions. The Commission will build on the experience of this first year. Regarding specifically the weaknesses identified in Italy, the ongoing conformity procedure will ensure that any damage to the EU financial interest will be corrected. Recommendation 3: the Commission actively monitors the application of remedial actions with regard to the deficiencies in the control system applicable to EU aid for producer groups in Poland (see paragraphs 3.36 to 3.38). X The Commission considers the recommendation as fully implemented. A remedial action plan is ongoing and is being closely monitored by the Commission. Delays in the implementation have triggered a suspension of the payments. An ongoing conformity procedure will ensure that the EU financial interest is protected. 2013 Following this review and the findings and conclusions for 2013, the Court recommends in the area of rural development that: Recommendation 1: the Member States carry out their existing administrative checks better, by using all relevant information available to the paying agencies, as this has the potential to detect and correct the majority of errors (see paragraphs 4.8 and 4.20). In particular for investment measures, administrative checks should use all available information to confirm the eligibility of the expenditure, project and beneficiary (including all ultimate shareholders) and compliance of public procurement procedures with the applicable EU and/or national rules; X The Commission discusses measures to improve the quality of administration and control systems with the Member States in meetings with the learning network, in DG AGRI’s simplification working groups, the Farm Advisory System (FAS), annual screening of LPIS quality and also in the conferences with the Directors of the PAs. The new legal framework for CAP 2014-2020 contains provisions which reduce risk of error. For rural development, various measures have been merged and streamlined, reducing the total number of measures. In this context, provisions have been clarified, eligibility rules have been modified in order to be easier to implement and contract conditions have been made more flexible. Simplified costs options, i.e. standard scales of cost units, lump sums and flat-rate financing can be applied so that the processes of claiming, administering and auditing reimbursement for payments made will be easier for everyone. Public procurement is one of the central elements in the action plans for rural development. Nevertheless the non-compliance of procurement does not necessarily mean that 100 % of the expenses involved are ineligible. The project as such may well achieve its goal and provide real added value. Moreover, it is noted that new guidelines for determining financial corrections to be made to expenditure financed by the Union under shared management, for non-compliance with the rules on public procurement, have been recently adopted (Commission Decision C(2013) 9527). 2013 Recommendation 2: the Commission ensures that all cases where the Court detected errors are followed up appropriately (as identified in paragraphs 4.14, 4.20 and 4.21, as well as in paragraph 4.16 for environment). X The Commission takes note that ECA’s analysis is on-going. Following the ECA remarks, in the latest updates of action plans the Commission and the ECA audit findings are more systematically followed-up. It should also be noted that all material DAS cases reported by the Court are followed up by the Commission, including through conformity procedure, if appropriate. The number of audits carried out by the Commission on EAFRD expenditure has also increased very substantially. National action plans were requested from the Member States and their implementation is being closely monitored by DG AGRI services. In the latest updates of the action plans, the Commission made sure that its own and ECA’s audit findings were more systematically followed up. Actions are time-framed and monitored through appropriate indicators. Geographical desks follow up the root causes of errors and mitigating actions in monitoring committees, annual review meetings with the managing authorities, bilateral meetings and programme amendments when relevant. Audit findings are communicated to geographical desks on a regular basis in order to allow them to discuss with Member States in view of improving the implementation of programmes. The Commission is organising through the European Network for Rural Development (ENRD) specific workshops on relevant issues aiming to reduce the error rate and improve the implementation of the policy. 2013 Recommendation 3: the Member States ensure that action plans to address the high error rate in rural development are complete, by including all regions and addressing all measures, particularly investment measures, and take the Commission's and Court's audit findings into account (see paragraph 4.30). X The Commission has already put in place a response to the high error rate. The Commission systematically recommends as part of the clearance procedure that Member States improve the control systems and ensure a more vigorous implementation thereof. Moreover, the Commission has asked systematically the paying agencies concerned by reservations to put in place action plans to remedy the shortcomings behind the reservations. When shortcomings are identified in the Member States’ management and control systems, action is taken at two levels: first through the opening of conformity clearance procedures to cover the financial risk for the EU budget; secondly, whenever necessary and appropriate, by asking the Member States to set up action plans to remedy the deficiencies. National action plans were requested from all Member States in January 2013 and their implementation is closely monitored by DG AGRI services (cf. the Commission staff working document on the assessment of root causes of errors in the implementation of rural development policy and corrective actions (SWD(2013) 244)). The Court has recognised the value of these measures and considers that they are a step in the right direction in addressing the causes of errors (see special report No 23/2014 on errors on rural development spending). In the latest updates of the action plans, the Commission made sure that its own and ECA’s audit findings were more systematically followed-up. In the recommendations made in special report No 23/2014, the Court stated that following preventive and corrective actions should be taken by the Member States where relevant. The Commission therefore does not consider that action plans have to be made for all regions and all measures but only where deficiencies have been identified. 2013 and for the CAP as a whole that: Recommendation 4: the Commission documents how it calculates the expenditure covered by its conformity audits (see paragraph 4.24). X The Commission will continue to report on the audit coverage in its AAR. Recommendation 5: the Commission takes steps to further reduce the backlog of open audit files, so as to enable all audits carried out prior to 2012 to be closed by the end of 2015 (see paragraph 4.27). X At the end of June 2016, only 20 files remained in the backlog. Recommendation 6: the Commission further develops its approach to calculating the RER by ensuring that it takes into account all expenditure and paying agencies (see paragraph 4.29). X The Commission considers the recommendation as fully implemented. The methodology for estimating the adjusted error rate is detailed in the Annex 4 to DG AGRI’s AAR. Moreover, following an IAS recommendation, detailed procedures on how the estimation of the remaining amount at risk has been drawn up (c.f. Ares(2014) 2509210 — 29/07/2014). These procedural steps were updated in 2015 and complemented by additional documentation on the conceptual model underlying the DG AGRI’s estimation of the error rate at the level of each paying agency. Furthermore, the Court recommends that: Recommendation 7: financial corrections to Member States concerning the European Fisheries Fund are supported by evidence of their validation (see paragraph 4.34). X 2012 Following this review and the findings and conclusions for 2012, the Court recommends that the Commission and Member States increase and speed up their efforts to ensure that: Recommendation 1: the eligibility of land, and in particular permanent pasture is properly recorded in the LPIS, especially in cases where areas are fully or partly covered with rocks, shrubs or dense trees or bushes or where land has been abandoned for several years. X The Commission is monitoring the situation: where weaknesses persist, action plans are being implemented, and conformity clearance procedures are ongoing. In its 2014 annual report, the Court did not address any recommendation to the Commission regarding IACS, and the same is true for the 2015 annual report. Recommendation 2: immediate remedial action is taken where administrative and control systems and/or IACS databases are found to be deficient or out of date. X When the Commission services detect such problems during the course of their audits, they request the Member State to take remedial actions. Where the problem is particularly acute, the Member State is required to implement a remedial action plan which is closely followed by the services. So far such plans have been found to be very effective. The Commission also ensures that the financial risk to the EU budget arising from such deficiencies is covered via financial corrections imposed via the conformity clearance procedure. Recommendation 3: payments are based on inspection results and that on-the-spot inspections are of the quality necessary to determine the eligible area in a reliable manner. X The Commission considers the recommendation as implemented in most respects .The Commission shares the view of the Court in this regard and it will continue to focus its audits on the risk of ineligible land being not detected prior to the payments. Financial corrections will be applied where necessary to protect the EU financial interests. 2012 Recommendation 4: the design and quality of the work performed by the directors of paying agencies and the certification bodies in support of their respective declarations and statements provide a reliable basis for the assessment of the legality and regularity of underlying transactions. X The Commission continues to review the work of the certification bodies via the analysis of the documents related to the financial clearance of accounts as well as audit missions to certification bodies. In addition, the guidelines necessary for the certification bodies to complete their work on the certification of the accounts for financial year 2015, which for the first time will include an assessment of the legality and regularity of the underlying transactions, were finalised in 2014 and the final version officially presented to Member States in January 2015. This included the updated guideline dealing with the revised management declaration. Since 2012 four expert group meetings took place where these guidelines were presented and discussed. In order to gain further assurance on the quality of the work of the CBs 15 missions took place in 2015 that will assess the CB’s work on legality and regularity, in particular focusing on the quality of their on-the-spot controls. Additional expert group meetings took place in 2015 and 2016 where experiences of the work on the legality and regularity of transactions were shared. Following this review and the findings and conclusions for 2012, the Court recommends in the area of rural development that: Recommendation 1: the Member States carry out their existing administrative checks better, by using all relevant information available to the paying agencies, as this has the potential to detect and correct the majority of errors. X See Commission’s reply to recommendation 1 for rural development from 2013. Recommendation 2: the Commission ensures that all cases where the Court detected errors are followed up appropriately. X 2012 Recommendation 3: the Commission, in DG AGRI's annual activity report, applies a similar approach for EAFRD as for decoupled area aid, where the Commission takes account of the results of its own conformity audits in assessing the error rate for each paying agency. X and for the CAP as a whole that: Recommendation 4: the Commission ensures adequate coverage in its conformity audits. X The Commission seeks to increase the audit coverage within the limits of the resources available. In the framework of its multiannual strategy, DG AGRI is applying from July 2014 a rolling three-year audit programme based on a central risk analysis and designed in a way to maximise the audit coverage. Recommendation 5: the Commission addresses the weaknesses identified in its conformity audits and the persistent problem of long delays in the conformity procedure as a whole. X The Commission will continue to improve its audit methodology and procedures. Moreover, it will continue to make efforts, notably in the framework of the implementation of the CAP reform, to improve and speed up the process bearing in mind the need to maintain quality standards and the Member State’s right of reply. Commission Implementing Regulation (EU) No 908/2014, applicable from January 2015, introduced compulsory deadlines for the different steps of the audit procedure. Reinforced monitoring tools have also been introduced. The backlog of files has almost been cleared by the end of June 2016. Recommendation 6: the Commission further improves its method of determining financial corrections so as to take better account of the nature and gravity of the infringements detected. X (1) the eligibility and size of agricultural parcels, in particular of permanent pasture, are correctly assessed and recorded by the Member States on the basis of the applicable EU criteria by way of comprehensive analysis of the most recent orthoimages (see paragraphs 3.9, 3.10 and 3.20); (2) immediate remedial action is taken by the Member States where the IACS is found to be affected by systemic errors especially as regards incorrect administrative treatment of aid applications (see paragraph 3.15 and Box 3.5); (3) the debtors ledgers of the Member States contain full and reliable information on the amounts and nature of debts and that effective recovery/enforcement procedures are applied without any undue delay (see paragraphs 3.24 and 3.25);

the eligibility and size of agricultural parcels, in particular of permanent pasture, are correctly assessed and recorded by the Member States on the basis of the applicable EU criteria by way of comprehensive analysis of the most recent orthoimages (see paragraphs 3.9, 3.10 and 3.20);

immediate remedial action is taken by the Member States where the IACS is found to be affected by systemic errors especially as regards incorrect administrative treatment of aid applications (see paragraph 3.15 and Box 3.5);

the debtors ledgers of the Member States contain full and reliable information on the amounts and nature of debts and that effective recovery/enforcement procedures are applied without any undue delay (see paragraphs 3.24 and 3.25);

CHAPTER 8 Global Europe and Security and citizenship

Txt. 1. THE COURT’S OBSERVATIONS INTRODUCTION This chapter has two parts. The first part presents our findings for Global Europe, and the second presents our findings for Security and citizenship. PART 1 — GLOBAL EUROPE Key information on the activities covered and spending in 2015 for Global Europe is provided in Figure 8.1. MFF heading 4 — Global Europe — Key information 2015 (billion euro) Source: 2015 consolidated accounts of the European Union. Total payments for the year 7,9 – advances 4,7 + clearings of advances 3,7 Audited population, total 6,9 THE COURT’S OBSERVATIONS Brief description of Global Europe Global Europe covers expenditure for all external action (foreign policy) funded by the EU budget. External policies are a major field of action for the EU, in particular in: (1) promoting and defending EU values abroad such as human rights, democracy and the rule of law; (2) projecting EU policies in support of addressing major global challenges such as combating climate change, reversing biodiversity loss, and protecting global public goods; (3) increasing the impact of EU development cooperation, with the primary aim of helping to eradicate poverty; (4) investing in the long-term prosperity and stability of the EU's neighbourhood both through preparing candidate countries for membership and through neighbourhood policy; (5) enhancing European solidarity following natural or man-made disasters; (6) improving crisis prevention and resolution, preserving peace and strengthening international security. Txt. 2. The most significant directorates-general involved in implementing the budget for external action are the Directorate-General for International Cooperation and Development (DG DEVCO), the Directorate-General for Neighbourhood and Enlargement Negotiations (DG NEAR), the Directorate-General for Humanitarian aid and Civil Protection (ECHO) and the Service for Foreign Policy Instruments (FPI). Payments for Global Europe in 2015 amount to 7,9 billion euro, financed under several instruments (see Figure 8.1) and delivery methods, and applied in more than 150 countries. Rules and procedures are often complex, including those for tendering and the award of contracts. Areas of expenditure less prone to errors In two areas the nature of the funding and the payment conditions limits the extent to which transactions are prone to errors. Those areas are budget support and the application of the so-called notional approach in multi-donor actions implemented by international organisations. Budget support Budget support contributes to a state’s general budget or its budget for a specific policy or objective. We examined whether the Commission had complied with specific conditions for making budget support payments to the partner country concerned and verified that the general eligibility conditions (such as satisfactory progress in public sector finance management) had been met. However, the legal provisions offer broad scope for interpretation and the Commission has considerable flexibility in deciding whether these general conditions have been met. Our audit of regularity cannot go beyond the stage at which aid is paid to the partner country. The funds transferred are then merged with the recipient country’s budget resources. Any weaknesses in its financial management leading to misuse at national level will not generate errors in our audit of regularity. Notional approach Where EU funding is subject to specific requirements regarding the eligibility of costs we assess whether those requirements are met. When the Commission’s contributions to multi-donor projects are pooled with those of other donors and are not earmarked for specific identifiable items of expenditure, the Commission assumes that EU eligibility rules are complied with as long as the pooled amount includes sufficient eligible expenditure to cover the EU contribution. This approach, as applied by the Commission, limits the work of the auditors when there are substantial contributions from other donors, which the Commission then considers sufficient to cover any ineligible expenditure we may have found. Audit scope and approach Annex 1.1 sets out the key elements of our audit approach and methodology. For the audit of Global Europe, the following specific points should be noted: (a) We examined a sample of 156 transactions as defined in Annex 1.1, paragraph 7. The sample is designed to be representative of the range of spending under the MFF heading. In 2015 the sample consisted of 64 transactions approved by the EU Delegations in 12 beneficiary states and 92 transactions approved by the Commission headquarters, of which 36 were implemented in humanitarian crisis areas through its ECHO partners. (b) Where we detected errors, we examined the relevant systems to identify specific system weaknesses. (c) We assessed the annual activity reports of DG DEVCO and DG NEAR. Txt. 3. Regularity of transactions Annex 8.1 contains a summary of the results of transaction testing. Of the 156 transactions examined, 40 (26 %) were affected by error. On the basis of the 30 errors we have quantified, the estimated level of error is 2,8 %. When excluding the seven budget support and 23 notional approach transactions, referred to in paragraphs 8.6 to 8.9, from the audited sample of 156 transactions, the estimated level of error is 3,8 %. In seven cases of quantifiable errors, the Commission had sufficient information to prevent, or detect and correct, the errors before accepting the expenditure. If all this information had been used to prevent, or detect and correct, errors, the estimated level of error would have been 1 percentage points lower. In nine other transactions, we found errors which should have been detected by the auditors appointed by the beneficiaries. These cases contribute 0,6 percentage points to the estimated level of error. Figure 8.2 presents the extent to which different types of error contributed to our estimated level of error for 2015. Global Europe — Breakdown of the estimated level of error Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The most frequent type of error (11 out of 30 cases), representing 32 % of the estimated level of error, concerns ineligible expenditure. This consists of expenditure related to activities not covered by a contract (see Figure 8.3) or incurred outside the eligibility period, non-compliance with the rule of origin, ineligible taxes and indirect costs wrongly charged as direct costs. Ineligible expenditure: activities not covered by a contract We examined expenditure declared to, and accepted by, the Commission for a contribution to an international organisation to reduce injecting drug use and its harmful consequences. The EU contribution amounted to 800000 euro. The international organisation implementing the action had claimed 16000 euro of costs relating to staff training. As the training was not covered by the contribution agreement, the costs are ineligible. Ineligible expenditure: activities not covered by a contract A pre-information letter for recovery has already been sent to the organisation. As for other costs, the Delegation will continue to ensure that implementing partners and project managers are aware that financial reports presented should meet all conditions stipulated by the relevant agreement. Concerning distribution of costs over the project period, particular attention will be paid where front-/end-loading is detected. 33 % of the estimated level of error represents six cases for which the Commission accepted (and in some cases cleared) expenditure not incurred at the moment the Commission accepted it (see Figure 8.4). Costs not actually incurred for the action Costs not actually incurred for the action We examined expenditure declared to, and accepted by, the Commission for a grant to a humanitarian organisation responding to the Syrian crisis. The EU grant amounted to 5800000 euro. Part of the costs of the action related to food transportation. The humanitarian organisation erroneously charged the EU for 8500 euro of food transportation for an action funded by another donor. We examined expenditure declared to, and accepted by, the Commission for a grant to support ballistic missile non-proliferation. The action with a budget of 930000 euro was financed in full by FPI. The beneficiary, a European-based research foundation, charged estimated staff costs that exceeded the actual incurred costs. FPI relies on expenditure verification reports before making a final payment for grants. Although the report for this contract had not identified this particular error, the Commission had already detected the error of the beneficiary through its system of ex post controls. FPI will also take measures to improve its current system for expenditure verification reports, following the example of DG DEVCO. The Commission is in the process of recovering the ineligible funds with a view to effectively protecting the financial interests of the EU. 14 % of the estimated level of error (eight cases) is due to expenditure for which essential supporting documentation was not provided. One case relates to the ineligible beneficiary category (see Figure 8.5). Ineligible beneficiary We examined expenditure declared to, and accepted by, the Education, Audiovisual and Culture Executive Agency (EACEA) to promote European cooperation in the youth field for an amount of 16500 euro. The grant was paid to a youth club, which we were unable to trace in order to perform an examination of the supporting documents. Ineligible beneficiary Following this finding, which suggests irregularity/fraudulent activities, the Commission has taken actions and followed the applicable rules and procedures and informed the relevant body (OLAF). The Commission has also taken necessary actions to recover the paid funds. 7 % of the estimated level of error arose from four cases where the beneficiary had not complied with the rules on procurement. In addition to the errors we have quantified, we identified 16 non-quantifiable errors relating to non-compliance with legal or contractual obligations. The seven budget support transactions examined contained no regularity errors (see also 2015 annual report on the European Development Funds, paragraph 38). Examination of annual activity reports and other elements of internal control Examination of annual activity reports Following a rotational approach in the area of Global Europe, we examined the 2015 annual activity report of DG NEAR. THE COURT’S OBSERVATIONS In his declaration of assurance for 2015, the Director-General of DG NEAR considers that, for both of the financial instruments managed by DG NEAR (ENI, IPA), the financial exposure from the amount at risk is below the materiality threshold of 2 % and the average determined error rate for the whole DG is 1,12 %. This statement is not consistent with our audit work. As this was the first year of DG NEAR operation, it has developed new methodologies for calculating error rates for the different programmes and management modes. DG NEAR recognises in its report that the approach applied needs further improvement. Our review of the different methodologies has been limited by some delays in the submission of information provided by DG NEAR. An in-depth review will be carried out for the 2016 annual activity report. DG NEAR calculated a residual error rate (RER) for 90 % of its expenditure resulting in three rates: RER for IPA direct management, RER for IPA indirect management with beneficiary countries and RER for ENI covering all management modes. For the remaining 10 % of expenditure DG NEAR used other sources of assurance. DG NEAR measured each of the three RERs on a different basis: Pt_a_. 2. RER for ENI (0,64 %) was determined by an external contractor following a methodology similar to the one applied for DG DEVCO RER. Pt_b_. 2. DG NEAR calculated RER for IPA under direct management (1,62 %) on the basis of its own ex post controls on a sample of transactions. Txt. 4. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Pt_c_. 2. DG NEAR relied for its RER methodology for IPA under indirect management with beneficiary countries on the audit authorities of the beneficiary countries. However, due to insufficient size of the populations, DG NEAR complemented the results with historical RERs for this management mode. The outcome was an error rate with a range of 0,02 — 1,84 %. Txt. 5. The management mode indirect management by beneficiary countries covers 37 % of payments made under IPA. We consider that the calculation of its RER, which combines results from non-statistical sampling by the audit authorities with the historical RER calculated by DG NEAR, is not sufficiently representative and does not provide sufficiently accurate information on the amount of payments at risk. There is a risk that the calculation, which is not fully in line with the Commission’s instructions, underestimates the level of error and may thus potentially impact on the assurance provided by the Director-General (e.g. the potential failure to make reservations). The countries working in the Indirect Management with Beneficiary Countries (IMBC) environment for which an RER was calculated, namely Croatia, FYROM and Turkey, have had a stable control environment for the past ten years. Hence the DG had no evidence of increased risk in this environment. The control framework has included 100 % ex ante controls on tendering and procurement and a very thorough clearance of accounts procedure on payments, along with regular on-the-spot checks by the Commission. In addition to this, the National Authorities, which have been entrusted by the Commission following a thorough entrustment audit process, have their own checks on both the contracting and payment phases of the process. The audit authorities in the countries perform further checks on the transactions of the national authorities. This engenders a low-risk environment and has led to an error rate below the materiality threshold for nine of the past ten years. 2015 was the first year that the opinion of the audit authorities of the countries was sought. As the error rate which the countries derived was considerably lower (0,02 %) than the historical average (below 2 %), and was based on non-statistical sampling, a central value (0,91 %) was elaborated based on the middle value between the 2015 AAR (0,02 %) and the highest RER (1,84 %) reported in the 2012 AAR and referring to the contracts under the 2005 programming year. This was designed to maintain the conservative and prudent approach of past years. This was based on Annex 2 of the DG BUDGET guidelines on the calculation of the RER, which explain the steps to be taken in situations such as the one described above. As a result the Commission does not consider that a reservation could have been required, as the materiality level finally arrived at was below 2 %. No residual error rate has been calculated for the management mode indirect management by entrusted entities for IPA covering 15 % of the payments under the instrument in 2015. The Commission points out that, although the RER rate for IPA in 2015 did not cover the IMEE management mode, it was covered by the following sources of assurance: (1) pillar assessments; (2) regular reporting by the entrusted entities concerned; (3) on-the-spot checks; (4) verification missions. The Commission would further stress that for 2015 under IPA IMEE, 94 % was made up of pre-financing payments. We have also examined DG DEVCO's 2015 annual activity report and the results are presented in our annual report on the 8th, 9th, 10th and 11th European Development Funds, paragraphs 49 to 53. System weaknesses revealed through transaction testing Transaction testing in the area of Global Europe has revealed some control weaknesses in the Commission’s systems. As stated in paragraph 8.13, the expenditure verification carried out by auditors appointed by beneficiaries in some cases failed to detect the errors leading to the Commission’s acceptance of ineligible costs. Our testing has also identified delays in the validation, authorisation and payment of expenditure by the Commission. DG DEVCO is going to apply a quality assurance system also for expenditure verifications directly contracted by the beneficiaries. However, since deficiencies in the system are partly linked to deficiencies in the Terms of Reference for these types of engagements, DG DEVCO will first revise these Terms of Reference. The revision of the Terms of Reference will include, in addition to reinforced quality requirements, a clearer focus on useful results for assurance purposes. Five of the transactions we examined relate to twinning contracts, all of which were financed under the European Neighbourhood and Partnership Instrument. Since twinning contracts are modelled on grant contracts, they are subject to the overall requirement that they may not yield a profit. We found that the specific rules set up by the Commission for the twinning instrument concerning lump sum and flat-rate costs do not define a maximum limit for funding, the purpose of which is mainly to cover the indirect costs of the Member State partner. For the audited contracts, the amount of flat-rate and lump-sum financing varied from 15 % to 27 % of the total costs. In comparison, the ceiling for indirect costs for grant contracts is generally set to 7 % and thus there is a risk that the implementing Member State partner yields a profit The Commission is preparing both an update of the Twining Manual and a new Commission Decision. The finding reported by the Court will be taken into consideration. DG DEVCO system assessment We have carried out an assessment of DG DEVCO systems and the detailed results of our work are presented in our annual report on the 8th, 9th, 10th and 11th European Development Funds, paragraphs 39 to 53. Conclusion and recommendations The conclusion for 2015 Overall audit evidence indicates that spending on Global Europe is affected by a material level of error. THE COURT’S OBSERVATIONS For this MFF heading area, testing of transactions indicates that the estimated level of error present in the population is 2,8 % (See Annex 8.1). Recommendations Annex 8.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, the Court presented seven recommendations. The Commission fully implemented five recommendations, while one was implemented in most respects and one in some respects. The results of our transaction testing have demonstrated, as in previous years, that the controls for detecting and preventing errors in grant contracts have in some instances failed. We recommended in our 2011 annual report that the Commission improve grant supervision and repeated the same recommendation in our 2014 annual report. We recognise that DG DEVCO is in the process of setting up new measures, such as systematic monitoring and on-the-spot visits, but as these actions are not yet implemented their impact can only be assessed in the years to come. As stated in paragraph 8.29, our testing has demonstrated that the quality of the expenditure verifications by the beneficiary-contracted auditors is not up to standard. We noted this weakness in our previous audits and recommended in our 2012 annual report that the Commission should take measures to enhance the quality of the expenditure verifications. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Following this review and the findings and conclusion for 2015, we recommend that: (13) Recommendation 1: DG DEVCO and DG NEAR enhance the quality of expenditure verifications contracted by beneficiaries, namely by introducing new measures such as the use of a quality grid to check the quality of the work performed by the beneficiary-contracted auditors and the revision of the auditors' terms of reference; Txt. 6. The Commission accepts recommendation 1. It is already developing an adaptation of the quality grid applied to audits and verifications conducted by the Commission, to the specific features, needs and implementation modalities of the expenditure verifications performed by the beneficiaries. (14) Recommendation 2: DG NEAR take action to ensure that funding channelled through a twinning instrument is in accordance with the non-profit rule and adheres to the principle of sound financial management; Txt. 7. The Commission accepts recommendation 2. (15) Recommendation 3: DG NEAR revise the RER methodology in order to provide statistically accurate information on the amount at risk for payments made under IPA indirect management. Txt. 8. The Commission accepts recommendation 3. PART 2 — SECURITY AND CITIZENSHIP Security and citizenship represents a small but increasing part of the EU budget. Key information on the activities covered and the spending in 2015 is provided in Figure 8.6. This shows a total audited population of 1,5 billion euro excluding decentralised agencies, or approximately 1 % of the EU budget. The present section focuses on the 0,8 billion euro spent on migration and security. MFF heading 3 — Security and citizenship — Key information 2015 (billion euro) Source: 2015 consolidated accounts of the European Union. Total payments for the year 2,0 – advances 1,3 + clearings of advances 1,4 Audited population, total 2,1 THE COURT’S OBSERVATIONS Brief description of Security and citizenship Security and citizenship groups different policies whose common objective is to strengthen the concept of European citizenship by creating an area of freedom, justice and security without internal frontiers. Expenditure is concentrated on three main policies: migration and security, food and feed, and creative Europe. Most expenditure for migration and security is from two funds: the Asylum, Migration and Integration Fund (AMIF), and the Internal Security Fund (ISF). The objective of AMIF is to contribute to the effective management of migration flows, as well as the implementation, strengthening and development of a common Union approach to asylum and immigration. The objective of the Internal Security Fund (ISF) is to contribute to a high level of security in the Union, by promoting the implementation of the Internal Security Strategy, law enforcement cooperation, and management of the Union’s external borders. The ISF is composed of two instruments, ISF Borders and Visa, and ISF Police. ISF Borders and Visa supports harmonised border management measures, and the development of a common visa policy. ISF Police supports cooperation between law enforcement agencies, and improving capacity to manage security related risks and crises. In 2014, AMIF and ISF replaced the previous SOLID programme (Solidarity and Management of Migration Flows). Like the SOLID programme, the management of most AMIF and ISF funds is shared between the Member States and the Commission. Audit scope and approach We performed a review of the main systems in the field of migration and security, under the responsibility of the Commission’s DG HOME. Concerning direct management, this included an examination of the Commission’s systems for grants and procurement. For shared management, this included an examination of the Commission’s assessment of Member State systems for managing SOLID funds. Summary of observations relating to DG HOME’s systems In 2015, payments amounting to 201 million euro under the migration and security policy area were directly managed by the Commission (DG HOME) through grants and procurement. The Commission implemented relevant controls required by the financial regulation. Our review did not reveal major weaknesses. In 2015, the Commission made payments amounting to 582 million euro for programmes under the shared management of the Commission and Member States. By the end of 2015, the Commission had closed 506 annual Member State SOLID programmes. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The Commission audited a selection of Member States’ management and control systems used for the SOLID programmes. Our examination of this work found that the Commission’s audits did not include tests of the effectiveness of internal controls over most key processes (selection and award procedures, project monitoring, payments and accounting), but were rather focused on the description of those processes, and walkthrough testing (designed to understand and document control procedures) — which risks to be not sufficient for concluding on control effectiveness. The Commission acknowledges that the audits did not cover tests of controls on most key processes. However, the Commission would like to reiterate that under shared management, the assurance on the control effectiveness is gradually built, not only on the system audits but through a system approach, starting with the inception phase, DG HOME operational monitoring visits, annual closure, ex post audits and, in line with the Single Audit Principle, through the audit work of the Member State, which must ensure over the whole programming period the audit coverage of all key processes, as requested by Article 17(2) of Decision No 456/2008/EC of 5 March 2008. Moreover, at the stage of the first MCS audit visits, some processes were not started yet and therefore were not auditable. To mitigate any risks it identifies from these system audits, the Commission performs follow up system audits, financial corrections, and ex post audits. These ex post audits are carried out after the Commission clears a Member State’s cost claim and may lead to the recovery of ineligible expenditure. The Commission focuses primarily on the minority of cases where it considers reasonable assurance regarding the legality and regularity of payments is not provided by Member State systems. However, due to the limited testing of controls identified in the previous paragraph, there is a risk that some annual programmes with ineffective control systems have been considered by the Commission to provide reasonable assurance and will not be the focus of the Commission’s ex post audits. By the time of our audit, the Commission had performed ex post audits on only 13 of closed programmes which it considered to provide reasonable assurance. This limited audit coverage reduces the reliability of the eligible expenditure declared to the Commission.

THE COURT’S OBSERVATIONS INTRODUCTION This chapter has two parts. The first part presents our findings for Global Europe, and the second presents our findings for Security and citizenship. PART 1 — GLOBAL EUROPE Key information on the activities covered and spending in 2015 for Global Europe is provided in Figure 8.1. MFF heading 4 — Global Europe — Key information 2015 (billion euro) Source: 2015 consolidated accounts of the European Union. Total payments for the year 7,9 – advances 4,7 + clearings of advances 3,7 Audited population, total 6,9 THE COURT’S OBSERVATIONS Brief description of Global Europe Global Europe covers expenditure for all external action (foreign policy) funded by the EU budget. External policies are a major field of action for the EU, in particular in:

promoting and defending EU values abroad such as human rights, democracy and the rule of law;

projecting EU policies in support of addressing major global challenges such as combating climate change, reversing biodiversity loss, and protecting global public goods;

increasing the impact of EU development cooperation, with the primary aim of helping to eradicate poverty;

investing in the long-term prosperity and stability of the EU's neighbourhood both through preparing candidate countries for membership and through neighbourhood policy;

enhancing European solidarity following natural or man-made disasters;

improving crisis prevention and resolution, preserving peace and strengthening international security.

The most significant directorates-general involved in implementing the budget for external action are the Directorate-General for International Cooperation and Development (DG DEVCO), the Directorate-General for Neighbourhood and Enlargement Negotiations (DG NEAR), the Directorate-General for Humanitarian aid and Civil Protection (ECHO) and the Service for Foreign Policy Instruments (FPI). Payments for Global Europe in 2015 amount to 7,9 billion euro, financed under several instruments (see Figure 8.1) and delivery methods, and applied in more than 150 countries. Rules and procedures are often complex, including those for tendering and the award of contracts. Areas of expenditure less prone to errors In two areas the nature of the funding and the payment conditions limits the extent to which transactions are prone to errors. Those areas are budget support and the application of the so-called notional approach in multi-donor actions implemented by international organisations. Budget support Budget support contributes to a state’s general budget or its budget for a specific policy or objective. We examined whether the Commission had complied with specific conditions for making budget support payments to the partner country concerned and verified that the general eligibility conditions (such as satisfactory progress in public sector finance management) had been met. However, the legal provisions offer broad scope for interpretation and the Commission has considerable flexibility in deciding whether these general conditions have been met. Our audit of regularity cannot go beyond the stage at which aid is paid to the partner country. The funds transferred are then merged with the recipient country’s budget resources. Any weaknesses in its financial management leading to misuse at national level will not generate errors in our audit of regularity. Notional approach Where EU funding is subject to specific requirements regarding the eligibility of costs we assess whether those requirements are met. When the Commission’s contributions to multi-donor projects are pooled with those of other donors and are not earmarked for specific identifiable items of expenditure, the Commission assumes that EU eligibility rules are complied with as long as the pooled amount includes sufficient eligible expenditure to cover the EU contribution. This approach, as applied by the Commission, limits the work of the auditors when there are substantial contributions from other donors, which the Commission then considers sufficient to cover any ineligible expenditure we may have found. Audit scope and approach Annex 1.1 sets out the key elements of our audit approach and methodology. For the audit of Global Europe, the following specific points should be noted:

(a) We examined a sample of 156 transactions as defined in Annex 1.1, paragraph 7. The sample is designed to be representative of the range of spending under the MFF heading. In 2015 the sample consisted of 64 transactions approved by the EU Delegations in 12 beneficiary states and 92 transactions approved by the Commission headquarters, of which 36 were implemented in humanitarian crisis areas through its ECHO partners.

(b) Where we detected errors, we examined the relevant systems to identify specific system weaknesses.

(c) We assessed the annual activity reports of DG DEVCO and DG NEAR.

Regularity of transactions Annex 8.1 contains a summary of the results of transaction testing. Of the 156 transactions examined, 40 (26 %) were affected by error. On the basis of the 30 errors we have quantified, the estimated level of error is 2,8 %. When excluding the seven budget support and 23 notional approach transactions, referred to in paragraphs 8.6 to 8.9, from the audited sample of 156 transactions, the estimated level of error is 3,8 %. In seven cases of quantifiable errors, the Commission had sufficient information to prevent, or detect and correct, the errors before accepting the expenditure. If all this information had been used to prevent, or detect and correct, errors, the estimated level of error would have been 1 percentage points lower. In nine other transactions, we found errors which should have been detected by the auditors appointed by the beneficiaries. These cases contribute 0,6 percentage points to the estimated level of error. Figure 8.2 presents the extent to which different types of error contributed to our estimated level of error for 2015. Global Europe — Breakdown of the estimated level of error Source: European Court of Auditors. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The most frequent type of error (11 out of 30 cases), representing 32 % of the estimated level of error, concerns ineligible expenditure. This consists of expenditure related to activities not covered by a contract (see Figure 8.3) or incurred outside the eligibility period, non-compliance with the rule of origin, ineligible taxes and indirect costs wrongly charged as direct costs. Ineligible expenditure: activities not covered by a contract We examined expenditure declared to, and accepted by, the Commission for a contribution to an international organisation to reduce injecting drug use and its harmful consequences. The EU contribution amounted to 800000 euro. The international organisation implementing the action had claimed 16000 euro of costs relating to staff training. As the training was not covered by the contribution agreement, the costs are ineligible. Ineligible expenditure: activities not covered by a contract A pre-information letter for recovery has already been sent to the organisation. As for other costs, the Delegation will continue to ensure that implementing partners and project managers are aware that financial reports presented should meet all conditions stipulated by the relevant agreement. Concerning distribution of costs over the project period, particular attention will be paid where front-/end-loading is detected. 33 % of the estimated level of error represents six cases for which the Commission accepted (and in some cases cleared) expenditure not incurred at the moment the Commission accepted it (see Figure 8.4). Costs not actually incurred for the action Costs not actually incurred for the action We examined expenditure declared to, and accepted by, the Commission for a grant to a humanitarian organisation responding to the Syrian crisis. The EU grant amounted to 5800000 euro. Part of the costs of the action related to food transportation. The humanitarian organisation erroneously charged the EU for 8500 euro of food transportation for an action funded by another donor. We examined expenditure declared to, and accepted by, the Commission for a grant to support ballistic missile non-proliferation. The action with a budget of 930000 euro was financed in full by FPI. The beneficiary, a European-based research foundation, charged estimated staff costs that exceeded the actual incurred costs. FPI relies on expenditure verification reports before making a final payment for grants. Although the report for this contract had not identified this particular error, the Commission had already detected the error of the beneficiary through its system of ex post controls. FPI will also take measures to improve its current system for expenditure verification reports, following the example of DG DEVCO. The Commission is in the process of recovering the ineligible funds with a view to effectively protecting the financial interests of the EU. 14 % of the estimated level of error (eight cases) is due to expenditure for which essential supporting documentation was not provided. One case relates to the ineligible beneficiary category (see Figure 8.5). Ineligible beneficiary We examined expenditure declared to, and accepted by, the Education, Audiovisual and Culture Executive Agency (EACEA) to promote European cooperation in the youth field for an amount of 16500 euro. The grant was paid to a youth club, which we were unable to trace in order to perform an examination of the supporting documents. Ineligible beneficiary Following this finding, which suggests irregularity/fraudulent activities, the Commission has taken actions and followed the applicable rules and procedures and informed the relevant body (OLAF). The Commission has also taken necessary actions to recover the paid funds. 7 % of the estimated level of error arose from four cases where the beneficiary had not complied with the rules on procurement. In addition to the errors we have quantified, we identified 16 non-quantifiable errors relating to non-compliance with legal or contractual obligations. The seven budget support transactions examined contained no regularity errors (see also 2015 annual report on the European Development Funds, paragraph 38). Examination of annual activity reports and other elements of internal control Examination of annual activity reports Following a rotational approach in the area of Global Europe, we examined the 2015 annual activity report of DG NEAR. THE COURT’S OBSERVATIONS In his declaration of assurance for 2015, the Director-General of DG NEAR considers that, for both of the financial instruments managed by DG NEAR (ENI, IPA), the financial exposure from the amount at risk is below the materiality threshold of 2 % and the average determined error rate for the whole DG is 1,12 %. This statement is not consistent with our audit work. As this was the first year of DG NEAR operation, it has developed new methodologies for calculating error rates for the different programmes and management modes. DG NEAR recognises in its report that the approach applied needs further improvement. Our review of the different methodologies has been limited by some delays in the submission of information provided by DG NEAR. An in-depth review will be carried out for the 2016 annual activity report. DG NEAR calculated a residual error rate (RER) for 90 % of its expenditure resulting in three rates: RER for IPA direct management, RER for IPA indirect management with beneficiary countries and RER for ENI covering all management modes. For the remaining 10 % of expenditure DG NEAR used other sources of assurance. DG NEAR measured each of the three RERs on a different basis:

(a) RER for ENI (0,64 %) was determined by an external contractor following a methodology similar to the one applied for DG DEVCO RER.

(b) DG NEAR calculated RER for IPA under direct management (1,62 %) on the basis of its own ex post controls on a sample of transactions.

THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES

(c) DG NEAR relied for its RER methodology for IPA under indirect management with beneficiary countries on the audit authorities of the beneficiary countries. However, due to insufficient size of the populations, DG NEAR complemented the results with historical RERs for this management mode. The outcome was an error rate with a range of 0,02 — 1,84 %.

The management mode indirect management by beneficiary countries covers 37 % of payments made under IPA. We consider that the calculation of its RER, which combines results from non-statistical sampling by the audit authorities with the historical RER calculated by DG NEAR, is not sufficiently representative and does not provide sufficiently accurate information on the amount of payments at risk. There is a risk that the calculation, which is not fully in line with the Commission’s instructions, underestimates the level of error and may thus potentially impact on the assurance provided by the Director-General (e.g. the potential failure to make reservations). The countries working in the Indirect Management with Beneficiary Countries (IMBC) environment for which an RER was calculated, namely Croatia, FYROM and Turkey, have had a stable control environment for the past ten years. Hence the DG had no evidence of increased risk in this environment. The control framework has included 100 % ex ante controls on tendering and procurement and a very thorough clearance of accounts procedure on payments, along with regular on-the-spot checks by the Commission. In addition to this, the National Authorities, which have been entrusted by the Commission following a thorough entrustment audit process, have their own checks on both the contracting and payment phases of the process. The audit authorities in the countries perform further checks on the transactions of the national authorities. This engenders a low-risk environment and has led to an error rate below the materiality threshold for nine of the past ten years. 2015 was the first year that the opinion of the audit authorities of the countries was sought. As the error rate which the countries derived was considerably lower (0,02 %) than the historical average (below 2 %), and was based on non-statistical sampling, a central value (0,91 %) was elaborated based on the middle value between the 2015 AAR (0,02 %) and the highest RER (1,84 %) reported in the 2012 AAR and referring to the contracts under the 2005 programming year. This was designed to maintain the conservative and prudent approach of past years. This was based on Annex 2 of the DG BUDGET guidelines on the calculation of the RER, which explain the steps to be taken in situations such as the one described above. As a result the Commission does not consider that a reservation could have been required, as the materiality level finally arrived at was below 2 %. No residual error rate has been calculated for the management mode indirect management by entrusted entities for IPA covering 15 % of the payments under the instrument in 2015. The Commission points out that, although the RER rate for IPA in 2015 did not cover the IMEE management mode, it was covered by the following sources of assurance: (1) pillar assessments; (2) regular reporting by the entrusted entities concerned; (3) on-the-spot checks; (4) verification missions. The Commission would further stress that for 2015 under IPA IMEE, 94 % was made up of pre-financing payments. We have also examined DG DEVCO's 2015 annual activity report and the results are presented in our annual report on the 8th, 9th, 10th and 11th European Development Funds, paragraphs 49 to 53. System weaknesses revealed through transaction testing Transaction testing in the area of Global Europe has revealed some control weaknesses in the Commission’s systems. As stated in paragraph 8.13, the expenditure verification carried out by auditors appointed by beneficiaries in some cases failed to detect the errors leading to the Commission’s acceptance of ineligible costs. Our testing has also identified delays in the validation, authorisation and payment of expenditure by the Commission. DG DEVCO is going to apply a quality assurance system also for expenditure verifications directly contracted by the beneficiaries. However, since deficiencies in the system are partly linked to deficiencies in the Terms of Reference for these types of engagements, DG DEVCO will first revise these Terms of Reference. The revision of the Terms of Reference will include, in addition to reinforced quality requirements, a clearer focus on useful results for assurance purposes. Five of the transactions we examined relate to twinning contracts, all of which were financed under the European Neighbourhood and Partnership Instrument. Since twinning contracts are modelled on grant contracts, they are subject to the overall requirement that they may not yield a profit. We found that the specific rules set up by the Commission for the twinning instrument concerning lump sum and flat-rate costs do not define a maximum limit for funding, the purpose of which is mainly to cover the indirect costs of the Member State partner. For the audited contracts, the amount of flat-rate and lump-sum financing varied from 15 % to 27 % of the total costs. In comparison, the ceiling for indirect costs for grant contracts is generally set to 7 % and thus there is a risk that the implementing Member State partner yields a profit The Commission is preparing both an update of the Twining Manual and a new Commission Decision. The finding reported by the Court will be taken into consideration. DG DEVCO system assessment We have carried out an assessment of DG DEVCO systems and the detailed results of our work are presented in our annual report on the 8th, 9th, 10th and 11th European Development Funds, paragraphs 39 to 53. Conclusion and recommendations The conclusion for 2015 Overall audit evidence indicates that spending on Global Europe is affected by a material level of error. THE COURT’S OBSERVATIONS For this MFF heading area, testing of transactions indicates that the estimated level of error present in the population is 2,8 % (See Annex 8.1). Recommendations Annex 8.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, the Court presented seven recommendations. The Commission fully implemented five recommendations, while one was implemented in most respects and one in some respects. The results of our transaction testing have demonstrated, as in previous years, that the controls for detecting and preventing errors in grant contracts have in some instances failed. We recommended in our 2011 annual report that the Commission improve grant supervision and repeated the same recommendation in our 2014 annual report. We recognise that DG DEVCO is in the process of setting up new measures, such as systematic monitoring and on-the-spot visits, but as these actions are not yet implemented their impact can only be assessed in the years to come. As stated in paragraph 8.29, our testing has demonstrated that the quality of the expenditure verifications by the beneficiary-contracted auditors is not up to standard. We noted this weakness in our previous audits and recommended in our 2012 annual report that the Commission should take measures to enhance the quality of the expenditure verifications. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES Following this review and the findings and conclusion for 2015, we recommend that:

Recommendation 1: DG DEVCO and DG NEAR enhance the quality of expenditure verifications contracted by beneficiaries, namely by introducing new measures such as the use of a quality grid to check the quality of the work performed by the beneficiary-contracted auditors and the revision of the auditors' terms of reference;

The Commission accepts recommendation 1. It is already developing an adaptation of the quality grid applied to audits and verifications conducted by the Commission, to the specific features, needs and implementation modalities of the expenditure verifications performed by the beneficiaries.

Recommendation 2: DG NEAR take action to ensure that funding channelled through a twinning instrument is in accordance with the non-profit rule and adheres to the principle of sound financial management;

The Commission accepts recommendation 2.

Recommendation 3: DG NEAR revise the RER methodology in order to provide statistically accurate information on the amount at risk for payments made under IPA indirect management.

The Commission accepts recommendation 3. PART 2 — SECURITY AND CITIZENSHIP Security and citizenship represents a small but increasing part of the EU budget. Key information on the activities covered and the spending in 2015 is provided in Figure 8.6. This shows a total audited population of 1,5 billion euro excluding decentralised agencies, or approximately 1 % of the EU budget. The present section focuses on the 0,8 billion euro spent on migration and security. MFF heading 3 — Security and citizenship — Key information 2015 (billion euro) Source: 2015 consolidated accounts of the European Union. Total payments for the year 2,0 – advances 1,3 + clearings of advances 1,4 Audited population, total 2,1 THE COURT’S OBSERVATIONS Brief description of Security and citizenship Security and citizenship groups different policies whose common objective is to strengthen the concept of European citizenship by creating an area of freedom, justice and security without internal frontiers. Expenditure is concentrated on three main policies: migration and security, food and feed, and creative Europe. Most expenditure for migration and security is from two funds: the Asylum, Migration and Integration Fund (AMIF), and the Internal Security Fund (ISF). The objective of AMIF is to contribute to the effective management of migration flows, as well as the implementation, strengthening and development of a common Union approach to asylum and immigration. The objective of the Internal Security Fund (ISF) is to contribute to a high level of security in the Union, by promoting the implementation of the Internal Security Strategy, law enforcement cooperation, and management of the Union’s external borders. The ISF is composed of two instruments, ISF Borders and Visa, and ISF Police. ISF Borders and Visa supports harmonised border management measures, and the development of a common visa policy. ISF Police supports cooperation between law enforcement agencies, and improving capacity to manage security related risks and crises. In 2014, AMIF and ISF replaced the previous SOLID programme (Solidarity and Management of Migration Flows). Like the SOLID programme, the management of most AMIF and ISF funds is shared between the Member States and the Commission. Audit scope and approach We performed a review of the main systems in the field of migration and security, under the responsibility of the Commission’s DG HOME. Concerning direct management, this included an examination of the Commission’s systems for grants and procurement. For shared management, this included an examination of the Commission’s assessment of Member State systems for managing SOLID funds. Summary of observations relating to DG HOME’s systems In 2015, payments amounting to 201 million euro under the migration and security policy area were directly managed by the Commission (DG HOME) through grants and procurement. The Commission implemented relevant controls required by the financial regulation. Our review did not reveal major weaknesses. In 2015, the Commission made payments amounting to 582 million euro for programmes under the shared management of the Commission and Member States. By the end of 2015, the Commission had closed 506 annual Member State SOLID programmes. THE COURT’S OBSERVATIONS THE COMMISSION’S REPLIES The Commission audited a selection of Member States’ management and control systems used for the SOLID programmes. Our examination of this work found that the Commission’s audits did not include tests of the effectiveness of internal controls over most key processes (selection and award procedures, project monitoring, payments and accounting), but were rather focused on the description of those processes, and walkthrough testing (designed to understand and document control procedures) — which risks to be not sufficient for concluding on control effectiveness. The Commission acknowledges that the audits did not cover tests of controls on most key processes. However, the Commission would like to reiterate that under shared management, the assurance on the control effectiveness is gradually built, not only on the system audits but through a system approach, starting with the inception phase, DG HOME operational monitoring visits, annual closure, ex post audits and, in line with the Single Audit Principle, through the audit work of the Member State, which must ensure over the whole programming period the audit coverage of all key processes, as requested by Article 17(2) of Decision No 456/2008/EC of 5 March 2008. Moreover, at the stage of the first MCS audit visits, some processes were not started yet and therefore were not auditable. To mitigate any risks it identifies from these system audits, the Commission performs follow up system audits, financial corrections, and ex post audits. These ex post audits are carried out after the Commission clears a Member State’s cost claim and may lead to the recovery of ineligible expenditure. The Commission focuses primarily on the minority of cases where it considers reasonable assurance regarding the legality and regularity of payments is not provided by Member State systems. However, due to the limited testing of controls identified in the previous paragraph, there is a risk that some annual programmes with ineffective control systems have been considered by the Commission to provide reasonable assurance and will not be the focus of the Commission’s ex post audits. By the time of our audit, the Commission had performed ex post audits on only 13 of closed programmes which it considered to provide reasonable assurance. This limited audit coverage reduces the reliability of the eligible expenditure declared to the Commission.

ANNEX

Anx. 81. 2015 2014 SIZE AND STRUCTURE OF THE SAMPLE Total transactions: 156 172 ESTIMATED IMPACT OF QUANTIFIABLE ERRORS Estimated level of error 2,8 % 2,7 % Upper Error Limit (UEL) 4,6 % Lower Error Limit (LEL) 1,0 %

ANNEX 8.1 RESULTS OF TRANSACTION TESTING FOR GLOBAL EUROPE

2015 2014 SIZE AND STRUCTURE OF THE SAMPLE Total transactions: 156 172 ESTIMATED IMPACT OF QUANTIFIABLE ERRORS Estimated level of error 2,8 % 2,7 % Upper Error Limit (UEL) 4,6 % Lower Error Limit (LEL) 1,0 %

Anx. 82. Year Court recommendation Court's analysis of the progress made Commission reply Fully implemented Being implemented Not implemented Not applicable Insufficient evidence In most respects In some respects 2013 The Court recommends that: Recommendation 1: the Commission, and particularly DG ELARG, ensure that instructions to staff state that clearings should be made only on the basis of incurred expenditure and not be based on their own estimates; X Recommendation 2: FPI accredit all CFSP missions in accordance with the six-pillar assessment. X As of the time of publication of this report, there are 10 ongoing CSDP missions. Four missions are currently compliant with Article 60 of the Financial Regulation: EULEX Kosovo, EUPOL Afghanistan, EUMM Georgia and EUPOL COPPS. These missions represent a substantial part of the budget of the CFSP. Work is continuing on the compliance with Article 60 of the remaining missions, starting with EUCAP Sahel Niger and EUCAP Nestor. Moreover the Commission will launch the pillar assessment of three additional missions this year (EUBAM Rafah, EUAM Ukraine and EUCAP Mali). Regarding the mission which has not yet been assessed (EUBAM Libya), this is not currently possible due to the security situation. 2012 The Court recommends that the Commission should: Recommendation 1: ensure timely clearance of expenditure; X Recommendation 2: promote better document management by implementing partners and beneficiaries; X Recommendation 3: improve the management of contract awarding procedures by setting out clear selection criteria and documenting the evaluation process better; X Recommendation 4: take effective steps in order to enhance the quality of expenditure checks carried out by external auditors; X Recommendation 5: apply a consistent and robust methodology for the external relations directorates-general to calculate residual error rates. X

ANNEX 8.2 FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR GLOBAL EUROPE

Year Court recommendation Court's analysis of the progress made Commission reply Fully implemented Being implemented Not implemented Not applicable Insufficient evidence In most respects In some respects 2013 The Court recommends that: Recommendation 1: the Commission, and particularly DG ELARG, ensure that instructions to staff state that clearings should be made only on the basis of incurred expenditure and not be based on their own estimates; X Recommendation 2: FPI accredit all CFSP missions in accordance with the six-pillar assessment. X As of the time of publication of this report, there are 10 ongoing CSDP missions. Four missions are currently compliant with Article 60 of the Financial Regulation: EULEX Kosovo, EUPOL Afghanistan, EUMM Georgia and EUPOL COPPS. These missions represent a substantial part of the budget of the CFSP. Work is continuing on the compliance with Article 60 of the remaining missions, starting with EUCAP Sahel Niger and EUCAP Nestor. Moreover the Commission will launch the pillar assessment of three additional missions this year (EUBAM Rafah, EUAM Ukraine and EUCAP Mali). Regarding the mission which has not yet been assessed (EUBAM Libya), this is not currently possible due to the security situation. 2012 The Court recommends that the Commission should: Recommendation 1: ensure timely clearance of expenditure; X Recommendation 2: promote better document management by implementing partners and beneficiaries; X Recommendation 3: improve the management of contract awarding procedures by setting out clear selection criteria and documenting the evaluation process better; X Recommendation 4: take effective steps in order to enhance the quality of expenditure checks carried out by external auditors; X Recommendation 5: apply a consistent and robust methodology for the external relations directorates-general to calculate residual error rates. X

CHAPTER 9 Administration

Txt. 1. THE COURT’S OBSERVATIONS INTRODUCTION This chapter presents our findings for Administration. Key information on the activities covered and the spending in 2015 is provided in Figure 9.1. MFF heading 5 — Administration — Key information 2015 Source: 2015 consolidated accounts of the European Union. (billion euro) Total payments for the year 9,0 - advances 0,1 + clearings of advances 0,1 Audited population, total 9,0 We report separately on the EU agencies and other bodies. Our mandate does not cover the financial audit of the European Central Bank. THE COURT’S OBSERVATIONS Brief description of the MFF heading area Administrative expenditure comprises expenditure on human resources, which accounts for about 60 % of the total, as well as expenditure on buildings, equipment, energy, communications and information technology. Overall, we consider this MFF heading to be a low-risk area. The main risks regarding administrative expenditure are non-compliance with the procedures for procurement, the implementation of contracts for services, supplies and works, recruitment and the calculation of salaries, pensions and allowances. Audit scope and approach Annex 1.1 sets out the key elements of our approach and methodology. For the audit of Administration, the following specific points should be noted: (a) We examined a sample of 151 transactions as defined in Annex 1.1, paragraph 7. The sample is designed to be representative of the range of spending under the MFF heading (see Figure 9.1 and paragraph 9.3). (b) We also examined procedures for recruiting temporary, contract and local staff and procurement procedures at the European External Action Service (EEAS) and the European Data Protection Supervisor. (c) We assessed the annual activity reports (AARs) of the European Commission’s directorates-general (DGs) and offices primarily responsible for administrative expenditure and of all the other institutions and bodies. Txt. 2. The European Court of Auditors’ own spending is audited by an external firm. The results of its audit of our financial statements for the year ending 31 December 2015 are presented in paragraph 9.15. REGULARITY OF TRANSACTIONS Annex 9.1 contains a summary of the results of transaction testing. Of the 151 transactions examined, 22 (14,6 %) were affected by error. On the basis of the seven errors we have quantified, the estimated level of error is 0,6 %. Figure 9.2 presents the extent to which different types of errors contributed to our estimated level of error for 2015. Administration — Breakdown of the estimated level of error Source: European Court of Auditors. THE COURT’S OBSERVATIONS EXAMINATION OF ANNUAL ACTIVITY REPORTS AND OTHER ELEMENTS OF INTERNAL CONTROL SYSTEMS We did not identify any serious weaknesses in the examined AARs and internal control systems of the institutions and bodies. A summary of our observations is presented in paragraphs 9.10 to 9.15. Observations on specific institutions and bodies We did not identify any serious weaknesses in respect of the issues examined for the Council, the European Court of Justice, the European Economic and Social Committee, the Committee of the Regions, the European Ombudsman or the European Data Protection Supervisor. The observations reported below have led to recommendations (see paragraph 9.18). European Parliament REPLY OF THE EUROPEAN PARLIAMENT Of the 16 European Parliament transactions we examined, one related to a political group. We found weaknesses in controls of the authorisation and settlement of expenditure made in 2014 but cleared in 2015, and in the related procurement procedures. The Parliament acknowledges the observation of the Court and has informed accordingly the political group concerned. Indeed, according to Article 1 of the rules on the use of appropriations from budget Item 4 0 0, adopted by the Bureau on 30 June 2003, the political groups manage the funds allocated to them according to the principles of indirect management of funds in analogical application of Article 60 of the Financial Regulation, whereby the specific requirements of the groups are taken into account in accordance with the provisions laid down in this regulation. European Commission THE COMMISSION’S REPLY As in previous years, we detected some weaknesses in the management of family allowances, and a small number of errors relating to the calculation of staff costs by the PMO. Measures have been taken to correct the detected weaknesses. As from 2016, all household allowances granted on the basis of spouse’s income will have an end date ≤ n+4 (n being the latest exercise carried out by the PMO to inquire about the spouse's revenues). This will result in a default verification before expiry of the recovery period established under Article 85 of the Staff Regulations. Additionally, the proactive notifications system in Sysper is being developed according to planning (2016), to facilitate the management of family allowances via notifications sent to the staff concerned. Strengthening of the internal control system (e.g. through targeted training) has further mitigated the risk of weaknesses in staff cost calculations. European External Action Service REPLY OF THE EUROPEAN EXTERNAL ACTION SERVICE We found weaknesses in recruitment procedures of local agents in delegations. Three of the six cases we examined were affected by a lack of transparency due to insufficient evidence that certain steps in the procedure had been completed. While the EEAS believes that all legal requirements have been met, it is also recognised that enhanced documentation would strengthen the transparency of the process. For that reason, we have clarified the recruitment procedures as prescribed in the delegations’ guide, in particular with respect to the shortlisting phase. REPLY OF THE EUROPEAN EXTERNAL ACTION SERVICE We also found weaknesses in procurement procedures organised by delegations for contracts worth less than 60000 euro. In five of the 12 cases examined, the delegations did not properly assess whether the tenders complied with the technical specifications, or they issued tender documents containing incomplete, incorrect or inconsistent information. In one other case, the use of an exceptional negotiated procedure with a single tenderer was not justified. The EEAS welcomes that no weaknesses were identified in the eight procurement procedures for contracts worth more than 60000 euros which are subject to Headquarters ex-ante controls. As for contracts worth less than 60000 euro organised by delegations which are concerned by this finding, measures are being implemented aiming at improving the quality of these procedures already as from 2016. In this respect, it is worth mentioning in particular the pilot project in the framework of the reform of administrative support to delegations and efforts to provide more training opportunities and ad-hoc support in this critical domain to delegations. European Court of Auditors The external auditor’s report states that the financial statements give a true and fair view of the financial position of the European Court of Auditors as of 31 December 2015, and of its financial performance, its cash flows and the changes in net assets for the year then ended. CONCLUSION AND RECOMMENDATIONS The conclusion for 2015 Overall audit evidence indicates that spending on Administration is not affected by a material level of error. For this MFF heading area testing of transactions indicates that the estimated level of error present in the population is 0,6 % (see Annex 9.1). Recommendations Annex 9.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, we presented five recommendations. Of these recommendations, two were not reviewed this year given our rotational approach. The institutions and bodies concerned implemented one recommendation in most respects and two in some respects. Following this review and the findings and conclusions for 2015, we recommend that: REPLY OF THE EUROPEAN PARLIAMENT (4) Recommendation 1: the European Parliament review the existing control framework for the implementation of budget appropriations allocated to political groups. In addition, the European Parliament provide better guidance through reinforced monitoring on the application by the groups of the rules for authorisation and settlement of expenditure, and for procurement procedures (see paragraph 9.11); Txt. 3. Parliament takes the view that it is primarily the responsibility of the political group concerned to take the necessary actions in order to remove the prevailing weaknesses. In this context, the political group concerned has set up a working group with the aim of identifying possible improvements in the financial management and the internal legal framework relating to the finances of the group. Parliament’s Secretariat will also extend the communication with regard to its existing guidance to political groups, through trainings, including tailor-made trainings and an explanatory note on procurement resuming the main elements to be respected. Taking into account the general task assigned to the external auditors to verify the accounts of the political groups, the groups will be asked to take the necessary initiatives to implement — with the assistance of their external auditors — efficient internal control systems. THE COMMISSION’S REPLY (5) Recommendation 2: the European Commission improve its monitoring systems with a view to the timely updating of the personal situation of staff members which may have an impact on the calculation of family allowances (see paragraph 9.12); Txt. 4. The Commission accepts the recommendation and has already taken measures to improve its monitoring systems in place, as detailed under point 9.12 above. REPLY OF THE EUROPEAN EXTERNAL ACTION SERVICE (6) Recommendation 3: the EEAS ensure that all steps in procedures to select local agents recruited in delegations are properly documented (see paragraph 9.13); Txt. 5. The EEAS welcomes this recommendation. The EU delegations’ guide has been amended to provide more detailed guidance on recruitment procedures, including the requirement of documenting properly the shortlisting process. REPLY OF THE EUROPEAN EXTERNAL ACTION SERVICE (7) Recommendation 4: the EEAS improve its guidance on the design, coordination and execution of procurement procedures in delegations for contracts worth less than 60000 euros (see paragraph 9.14). Txt. 6. The EEAS is implementing measures aiming at improving the quality of procedures applicable to contracts worth less than 60000 euros in delegations already as from 2016. In this respect, it is worth mentioning in particular the pilot project in the framework of the reform of administrative support to delegations and efforts to provide more training opportunities and ad-hoc support m this critical domain to delegations

THE COURT’S OBSERVATIONS INTRODUCTION This chapter presents our findings for Administration. Key information on the activities covered and the spending in 2015 is provided in Figure 9.1. MFF heading 5 — Administration — Key information 2015 Source: 2015 consolidated accounts of the European Union. (billion euro) Total payments for the year 9,0 - advances 0,1 + clearings of advances 0,1 Audited population, total 9,0 We report separately on the EU agencies and other bodies. Our mandate does not cover the financial audit of the European Central Bank. THE COURT’S OBSERVATIONS Brief description of the MFF heading area Administrative expenditure comprises expenditure on human resources, which accounts for about 60 % of the total, as well as expenditure on buildings, equipment, energy, communications and information technology. Overall, we consider this MFF heading to be a low-risk area. The main risks regarding administrative expenditure are non-compliance with the procedures for procurement, the implementation of contracts for services, supplies and works, recruitment and the calculation of salaries, pensions and allowances. Audit scope and approach Annex 1.1 sets out the key elements of our approach and methodology. For the audit of Administration, the following specific points should be noted:

(a) We examined a sample of 151 transactions as defined in Annex 1.1, paragraph 7. The sample is designed to be representative of the range of spending under the MFF heading (see Figure 9.1 and paragraph 9.3).

(b) We also examined procedures for recruiting temporary, contract and local staff and procurement procedures at the European External Action Service (EEAS) and the European Data Protection Supervisor.

(c) We assessed the annual activity reports (AARs) of the European Commission’s directorates-general (DGs) and offices primarily responsible for administrative expenditure and of all the other institutions and bodies.

The European Court of Auditors’ own spending is audited by an external firm. The results of its audit of our financial statements for the year ending 31 December 2015 are presented in paragraph 9.15. REGULARITY OF TRANSACTIONS Annex 9.1 contains a summary of the results of transaction testing. Of the 151 transactions examined, 22 (14,6 %) were affected by error. On the basis of the seven errors we have quantified, the estimated level of error is 0,6 %. Figure 9.2 presents the extent to which different types of errors contributed to our estimated level of error for 2015. Administration — Breakdown of the estimated level of error Source: European Court of Auditors. THE COURT’S OBSERVATIONS EXAMINATION OF ANNUAL ACTIVITY REPORTS AND OTHER ELEMENTS OF INTERNAL CONTROL SYSTEMS We did not identify any serious weaknesses in the examined AARs and internal control systems of the institutions and bodies. A summary of our observations is presented in paragraphs 9.10 to 9.15. Observations on specific institutions and bodies We did not identify any serious weaknesses in respect of the issues examined for the Council, the European Court of Justice, the European Economic and Social Committee, the Committee of the Regions, the European Ombudsman or the European Data Protection Supervisor. The observations reported below have led to recommendations (see paragraph 9.18). European Parliament REPLY OF THE EUROPEAN PARLIAMENT Of the 16 European Parliament transactions we examined, one related to a political group. We found weaknesses in controls of the authorisation and settlement of expenditure made in 2014 but cleared in 2015, and in the related procurement procedures. The Parliament acknowledges the observation of the Court and has informed accordingly the political group concerned. Indeed, according to Article 1 of the rules on the use of appropriations from budget Item 4 0 0, adopted by the Bureau on 30 June 2003, the political groups manage the funds allocated to them according to the principles of indirect management of funds in analogical application of Article 60 of the Financial Regulation, whereby the specific requirements of the groups are taken into account in accordance with the provisions laid down in this regulation. European Commission THE COMMISSION’S REPLY As in previous years, we detected some weaknesses in the management of family allowances, and a small number of errors relating to the calculation of staff costs by the PMO. Measures have been taken to correct the detected weaknesses. As from 2016, all household allowances granted on the basis of spouse’s income will have an end date ≤ n+4 (n being the latest exercise carried out by the PMO to inquire about the spouse's revenues). This will result in a default verification before expiry of the recovery period established under Article 85 of the Staff Regulations. Additionally, the proactive notifications system in Sysper is being developed according to planning (2016), to facilitate the management of family allowances via notifications sent to the staff concerned. Strengthening of the internal control system (e.g. through targeted training) has further mitigated the risk of weaknesses in staff cost calculations. European External Action Service REPLY OF THE EUROPEAN EXTERNAL ACTION SERVICE We found weaknesses in recruitment procedures of local agents in delegations. Three of the six cases we examined were affected by a lack of transparency due to insufficient evidence that certain steps in the procedure had been completed. While the EEAS believes that all legal requirements have been met, it is also recognised that enhanced documentation would strengthen the transparency of the process. For that reason, we have clarified the recruitment procedures as prescribed in the delegations’ guide, in particular with respect to the shortlisting phase. REPLY OF THE EUROPEAN EXTERNAL ACTION SERVICE We also found weaknesses in procurement procedures organised by delegations for contracts worth less than 60000 euro. In five of the 12 cases examined, the delegations did not properly assess whether the tenders complied with the technical specifications, or they issued tender documents containing incomplete, incorrect or inconsistent information. In one other case, the use of an exceptional negotiated procedure with a single tenderer was not justified. The EEAS welcomes that no weaknesses were identified in the eight procurement procedures for contracts worth more than 60000 euros which are subject to Headquarters ex-ante controls. As for contracts worth less than 60000 euro organised by delegations which are concerned by this finding, measures are being implemented aiming at improving the quality of these procedures already as from 2016. In this respect, it is worth mentioning in particular the pilot project in the framework of the reform of administrative support to delegations and efforts to provide more training opportunities and ad-hoc support in this critical domain to delegations. European Court of Auditors The external auditor’s report states that the financial statements give a true and fair view of the financial position of the European Court of Auditors as of 31 December 2015, and of its financial performance, its cash flows and the changes in net assets for the year then ended. CONCLUSION AND RECOMMENDATIONS The conclusion for 2015 Overall audit evidence indicates that spending on Administration is not affected by a material level of error. For this MFF heading area testing of transactions indicates that the estimated level of error present in the population is 0,6 % (see Annex 9.1). Recommendations Annex 9.2 shows the result of our review of progress in addressing recommendations made in previous annual reports. In the 2012 and 2013 annual reports, we presented five recommendations. Of these recommendations, two were not reviewed this year given our rotational approach. The institutions and bodies concerned implemented one recommendation in most respects and two in some respects. Following this review and the findings and conclusions for 2015, we recommend that: REPLY OF THE EUROPEAN PARLIAMENT

Recommendation 1: the European Parliament review the existing control framework for the implementation of budget appropriations allocated to political groups. In addition, the European Parliament provide better guidance through reinforced monitoring on the application by the groups of the rules for authorisation and settlement of expenditure, and for procurement procedures (see paragraph 9.11);

Parliament takes the view that it is primarily the responsibility of the political group concerned to take the necessary actions in order to remove the prevailing weaknesses. In this context, the political group concerned has set up a working group with the aim of identifying possible improvements in the financial management and the internal legal framework relating to the finances of the group. Parliament’s Secretariat will also extend the communication with regard to its existing guidance to political groups, through trainings, including tailor-made trainings and an explanatory note on procurement resuming the main elements to be respected. Taking into account the general task assigned to the external auditors to verify the accounts of the political groups, the groups will be asked to take the necessary initiatives to implement — with the assistance of their external auditors — efficient internal control systems. THE COMMISSION’S REPLY

Recommendation 2: the European Commission improve its monitoring systems with a view to the timely updating of the personal situation of staff members which may have an impact on the calculation of family allowances (see paragraph 9.12);

The Commission accepts the recommendation and has already taken measures to improve its monitoring systems in place, as detailed under point 9.12 above. REPLY OF THE EUROPEAN EXTERNAL ACTION SERVICE

Recommendation 3: the EEAS ensure that all steps in procedures to select local agents recruited in delegations are properly documented (see paragraph 9.13);

The EEAS welcomes this recommendation. The EU delegations’ guide has been amended to provide more detailed guidance on recruitment procedures, including the requirement of documenting properly the shortlisting process. REPLY OF THE EUROPEAN EXTERNAL ACTION SERVICE

Recommendation 4: the EEAS improve its guidance on the design, coordination and execution of procurement procedures in delegations for contracts worth less than 60000 euros (see paragraph 9.14).

The EEAS is implementing measures aiming at improving the quality of procedures applicable to contracts worth less than 60000 euros in delegations already as from 2016. In this respect, it is worth mentioning in particular the pilot project in the framework of the reform of administrative support to delegations and efforts to provide more training opportunities and ad-hoc support m this critical domain to delegations

ANNEX

Anx. 91. 2015 2014 SIZE AND STRUCTURE OF THE SAMPLE Total transactions: 151 129 ESTIMATED IMPACT OF QUANTIFIABLE ERRORS Estimated level of error 0,6 % 0,5 % Upper Error Limit (UEL) 1,2 % Lower Error Limit (LEL) 0,0 %

ANNEX 9.1 RESULTS OF TRANSACTION TESTING FOR ADMINISTRATION

2015 2014 SIZE AND STRUCTURE OF THE SAMPLE Total transactions: 151 129 ESTIMATED IMPACT OF QUANTIFIABLE ERRORS Estimated level of error 0,6 % 0,5 % Upper Error Limit (UEL) 1,2 % Lower Error Limit (LEL) 0,0 %

Anx. 92. Year Court recommendation Court's analysis of the progress made Institution's reply Fully implemented Being implemented Not implemented Not applicable Insufficient evidence In most respects In some respects 2013 Recommendation 1 (Commission): Updating of the personal situation and management of family allowances The Commission should take further steps to ensure that staff provide documents confirming their personal situation on a regular basis, as well as to improve systems for the timely processing of those documents that have an impact on the calculation of family allowances (see the 2013 annual report, paragraphs 9.11 and 9.20). X The Commission has taken measures to improve the monitoring system in place. In addition a systematic follow-up on the family situation of staff is being phased as from 2016. Please see reply to paragraph 9.12. Recommendation 2 (EEAS): Updating of the personal situation and management of family allowances The EEAS should take further steps to ensure that staff provide documents confirming their personal situation on a regular basis, as well as to improve systems for the timely processing of those documents that have an impact on the calculation of family allowances (see the 2013 annual report, paragraphs 9.13 and 9.20). X Recommendation 3 (EEAS): Procurement The EEAS should improve the design, coordination and conduct od procurement procedures by means of its headquarters providing increased support and guidance to the delegations (see the 2013 annual report, paragraphs 9.14, 9.15 and 9.20). X 2012 Recommendation 1 (European Parliament): Procurement The European Parliament should ensure that authorising officers improve the design, coordination and performance of procurement procedures through appropriate checks and better guidance (see the 2012 annual report, paragraphs 9.12 and 9.19). X Recommendation 2 (European Council and Council): Procurement The Council should ensure that authorising officers improve the design, coordination and performance of procurement procedures through appropriate checks and better guidance (see the 2012 annual report, paragraphs 9.14 and 9.19). X

ANNEX 9.2 FOLLOW-UP OF PREVIOUS RECOMMENDATIONS FOR ADMINISTRATION

Year Court recommendation Court's analysis of the progress made Institution's reply Fully implemented Being implemented Not implemented Not applicable Insufficient evidence In most respects In some respects 2013 Recommendation 1 (Commission): Updating of the personal situation and management of family allowances The Commission should take further steps to ensure that staff provide documents confirming their personal situation on a regular basis, as well as to improve systems for the timely processing of those documents that have an impact on the calculation of family allowances (see the 2013 annual report, paragraphs 9.11 and 9.20). X The Commission has taken measures to improve the monitoring system in place. In addition a systematic follow-up on the family situation of staff is being phased as from 2016. Please see reply to paragraph 9.12. Recommendation 2 (EEAS): Updating of the personal situation and management of family allowances The EEAS should take further steps to ensure that staff provide documents confirming their personal situation on a regular basis, as well as to improve systems for the timely processing of those documents that have an impact on the calculation of family allowances (see the 2013 annual report, paragraphs 9.13 and 9.20). X Recommendation 3 (EEAS): Procurement The EEAS should improve the design, coordination and conduct od procurement procedures by means of its headquarters providing increased support and guidance to the delegations (see the 2013 annual report, paragraphs 9.14, 9.15 and 9.20). X 2012 Recommendation 1 (European Parliament): Procurement The European Parliament should ensure that authorising officers improve the design, coordination and performance of procurement procedures through appropriate checks and better guidance (see the 2012 annual report, paragraphs 9.12 and 9.19). X Recommendation 2 (European Council and Council): Procurement The Council should ensure that authorising officers improve the design, coordination and performance of procurement procedures through appropriate checks and better guidance (see the 2012 annual report, paragraphs 9.14 and 9.19). X

Versions

1. 2016-10-13