Search papers

The EU Budget: Audit and Fraud


This paper discusses the EU’s finances with particular reference to two complaints frequently made about the European Union’s finances: that the European Court of Auditors each year ‘refuses to sign off the accounts’; and that the EU budget is subject to high levels of fraud.

€148.5 billion was spent from the EU budget in 2013: around one per cent of EU gross national income, equal to approximately two per cent of total public spend¬ing of EU Member States.1

The European Commission is responsible for implementing the budget, and for ensuring that EU funds are spent properly and in conformity with the relevant rules and regulations.2  The Commission exercises this responsibility in co-operation with the Member States, who are the day-to-day managers of about 80 per cent of EU expenditure, including spending on agriculture, rural development and ‘cohesion’/the structural funds, but who must satisfy the Commission that they properly manage the funds at their disposal.


Accounts and Audit

The Statement of Assurance

The European Court of Auditors (ECA) is the Union’s external auditor.  For each year since 1994 the Treaty has required it to ‘provide 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 underlying transactions’.3  In other words, do the accounts produced by the European institutions and consolidated by the Commission give ‘a true and fair view’; and has expenditure from the EU budget taken place (and revenue been received) as required by and in full conformity with the provisions of EU law?

The ECA publishes its statement of assurance each November, together with its annual report.  Much of the UK press habitually reports this event as the ECA’s ‘failing again to sign off the EU’s accounts’.


The Accounts

For every year since 2007 the ECA has given a positive, ‘clean’ opinion on the Union’s accounts (in the years before that the ECA qualified its approval of the accounts).  The statement of assurance for 2013, published in November 2014, says:

In the Court’s opinion, the consolidated accounts of the European Union for the year ended 31 December 2013 present fairly, in all material respects, the financial position of the Union as at 31 December 2013, 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.4

Although the ECA habitually includes in its annual reports a few observations on ways in which the accounts might be further developed and improved, the claim that the ECA ‘refuses to sign off the Commission’s accounts’ is incorrect and misleading.5


‘Underlying Transactions’

The position with the second part of the statement of assurance, dealing with the legality and regularity of the transactions that underlie the accounts, is more complex.  In 2013, as in previous years, the ECA found that revenue received by the EU was free from material error, as were commitments to make future payments.6

However the statement of assurance included an adverse opinion on the legality and regularity of payments that underlie the accounts: “In the Court’s opinion […] the payments underlying the accounts for the year ended 31 December 2013 are materially affected by error.”7

The ECA’s negative opinions on the legality and regularity of payments are based mainly on the results of its ‘transactions testing’.  It takes a representative random sample of budget payments in each broad area of expenditure (agricultural support, rural development, external spending, etc.) examines those transactions (over 1000 in total) in detail, all the way from the Commission’s budget accounts to the final recipients, to see if the relevant regulations have been correctly applied; identifies and where possible quantifies the results of ‘errors’ (failure correctly to apply the relevant regulations); and, using standard statistical techniques, extrapolates the results to give an estimate of the ‘most likely error rate’ (MLE) for each broad area of expenditure and for the budget as a whole.  It then compares the results with the figure of two per cent, chosen by the ECA as a ‘materiality threshold’.8

For 2013 the ECA estimated that MLE for payments from the budget as a whole was 4.7 per cent.  For all broad areas of operational expenditure MLE exceeded two per cent.


Interpreting the ECA’s negative opinion on payments

MLE is what it says: a most likely estimate.  On the other hand there is a reasonable degree of stability in the ECA’s audit results (for example MLE for payments from the budget as a whole in 2012 was 4.8 per cent).  The results should be taken seriously as an indication of shortcomings in the management of the EU budget.9

The ECA’s annual reports and, in respect of the largest categories of expenditure, its report Agriculture and cohesion: overview of EU spending 2007-13, published in 2014, provide a range of information about the management of the budget and the nature of the errors that the ECA identifies in its audits.10


Agriculture and cohesion

In its overview report the ECA states that, taking the years 2009-2013 as a whole, the most likely rate of error for agriculture was 3.7 per cent and for cohesion six per cent, and this relates overwhelmingly to errors occurring in Member States’ day-to-day administration of this expenditure.  Within agricultural spending, direct and market support was significantly less error prone than rural development spending.11  The ECA’s 2013 results are consistent with this pattern.  The MLE for agricultural market support was 3.6 per cent, compared with 7.9 per cent for rural development, environment and health, and 6.8 per cent for regional policy, transport and energy.12


What sort of ‘errors’?

In 2013, the types of error with the highest financial impact were as follows:

  • ineligible costs included in cost claims (contributing 1.8 percentage points of the 4.7 per cent MLE)
  • claims for ineligible activities, projects or beneficiaries (1.1 percentage points)
  • serious errors in public procurement (1.0 percentage points)
  • incorrect declaration of agricultural land (0.6 percentage points).13

The pattern is largely the same for agricultural and cohesion spending over the period 2009-13.14  Errors usually occur in the day-to-day administration of this ‘shared management’ expenditure by Member States.  The overview report identifies the main risks to regularity of agricultural spending as payments of subsidy on ineligible land, animals, or beneficiaries and the incorrect calculation of subsidies.  Breaches of agri environment requirements, of specific requirements for investment projects, and of procurement rules are important factors increasing the risk for rural development spending.  The biggest risk in cohesion spending relates to breaches of EU and/or national public procurement rules.  The next biggest risk is that expenditure is not eligible for EU subsidy.’15


Error or fraud?

The ECA does not attempt to quantify the extent to which fraud (deliberate deception) is responsible for the errors that it identifies, but in presenting its reports emphasises that MLE is an estimate of error – which can arise from genuine mistakes, both by claimants and by those managing payment schemes, in dealing with complex EU regulations – not necessarily of fraud or waste.  The ECA each year reports ‘several’ suspected cases of fraud to OLAF, the Union’s anti-fraud office (see below) – whereas there were several hundred cases in which the ECA identified some sort of error in 2013.16

The overview report concludes that a key challenge is to take action to make pro¬grammes easier to manage; and observes that the frequency with which the ECA finds errors illustrates among other things the complexity of EU spending programmes.17



The ECA does not attempt to estimate rates of error in individual Member States.  However in its overview report the ECA explains that error in cohesion and agricultural spending is geographically widespread and that it finds errors across the Union.  All eleven Member States (Poland, Spain, France, Germany, Italy, Greece, the United Kingdom, Portugal, Hungary, the Czech Republic and Romania) for which the ECA sampled sufficient transactions over the period to enable it to draw a valid conclusion were affected by material levels of error.  In each of the eleven, at least one third of the transactions examined involved some breach of legal requirements and at least one seventh of the transactions examined involved overpayments and similar, quantifiable, errors.18


Financial corrections and recoveries

The European Commission has extensive powers of audit and supervision over expenditure programmes, including in respect of agriculture and cohesion, where day-to-day management lies with Member States.  The Commission has in the past claimed that the ECA’s estimates of error are not the end of the story, since the Commission makes ‘financial corrections’ and recoveries in respect of expenditure programmes whose management it finds to be inadequate.  The amounts involved can be considerable: for example almost €3 billion in 2014 according to the Commission’s ‘Synthesis report’ on its management achievements for 2014.19  However to relate corrections and recoveries on the one hand and the ECA’s findings on the other is complicated.  Financial corrections and recoveries cannot simply be deducted from estimates of error.20



Negative opinions of a similar kind to those found in the accounts and budget of the EU also occur in the auditing of Member States’ own public accounts.  For example, the UK’s Comptroller and Auditor General has qualified the accounts of the UK Department of Work and Pensions and its predecessors for every year since 1988-89, due to the material level of fraud and error in benefit expenditure, (other than Retirement Pensions).21

The US Comptroller General has refused to give an opinion on the consolidated accounts of the US government in every year since 1996 (when such accounts were first prepared) because of serious financial management problems at the Department of Defense, the federal government’s inability adequately to account for and reconcile intra-governmental activity and balances between federal agencies, and the federal government’s ineffective process for preparing the consolidated financial statements.22


Fraud against the EU Budget

The EU has invested considerable effort to combat fraud against the EU budget.  Member States are under a Treaty obligation ‘to take the same measures to counter fraud affecting the financial interests of the Union as they take to counter fraud affecting their own financial interests’ and to co-operate with each other and the European Commission in combating fraud.23

The European Commission reports annually on cases of potential fraud identified by Member States and EU institutions and on action taken under a wide range of initiatives to minimise fraud against the budget.  In 2013, 976 cases of suspected fraud relating to EU budget expenditure were reported, with a total value of 248 million euros, equivalent to 0.19 per cent of EU expenditure the same year (together with 633 cases relating to the ‘traditional own resources’ – customs duties and levies – which finance part of the budget, equal to 0.29 per cent of total traditional own resources established in 2013).  As in previous years, expenditure on ‘cohesion’/the structural funds accounted for the largest proportion of reported fraudulent irregularities.24

These figures should be treated with care.  By its very nature, fraud is hard to quantify: fraudsters do not advertise their activities.  Whether a transaction is fraudulent is ultimately a decision for judicial authorities.  And the Commission point out that Member States have different approaches to dealing with fraud: some allocate significant resources to countering fraud; others prefer to apply financial corrections without further investigation of the potential criminal offence.  (In 2013, Italy, Romania, Bulgaria, Poland, Denmark and Greece reported the most cases of suspected fraud; the largest amounts were reported by Italy, Poland, Romania, Greece and Germany).


The European Anti-Fraud Office

The European Anti-Fraud Office or OLAF (derived from the French, Office européen de lutte antifraud) carries out independent investigations into fraud and corruption involving EU funds; and investigates cases of serious misconduct within EU institutions.  Its 420 officials are EU employees, but their investigations are independent.25

OLAF has extensive powers of investigation and can carry out on-the-spot checks on business premises in Member States and in some other countries as well (for example, where overseas aid funds are based).  OLAF officials work closely alongside the authorities of Member States.  In criminal investigations Member State prosecutors must take the lead.  OLAF cannot themselves bring cases to court.  When an investigation suggests that prosecution is justified, OLAF sends the file to the relevant national authority for action.

OLAF carries out investigations if it believes that there is sufficient suspicion of fraud, corruption or any other illegal activity affecting the EU’s financial interests or of serious misconduct by EU staff and members of the EU institutions in the discharge of their professional duties.  In 2014 OLAF began 234 investigations into alleged fraud against the budget and a further 54 ‘coordination cases’, in which OLAF assists and contributes to investigations carried out by Member States; closed 307 such investigations; carried out 40 investigations into alleged serious misconduct within the EU institutions; and made 397 recommendations for further action by EU institutions and bodies or Member States, including the recovery of €901 million (more than twice the average in the preceding four years).  Following the completion of significant inquiries into the structural funds, external aid and the customs and trade sectors (the structural funds accounted for just over half of the total), €206 million were recovered.26

OLAF’s work in 2014 included:

  • co-ordinating a successful international investigation into avoiding taxes by tobacco smuggling;
  • identifying deliberate and systematic over-claiming by a company that received funds under an EU scheme to divert the carriage of goods from road to other modes;
  • investigating a complex fraud under which the price of medical equipment funded from the European Regional Development Fund was artificially inflated, and some of the equipment either left unused or used in areas not eligible for support from the Fund; and
  • examining a case of potential corruption and conflict of interest by a member of staff in a Commission overseas delegation.27



Most debate about EU audit and fraud is polarised between gross exaggeration and complacent excuses.  Neither is a sensible or justifiable approach.  For example, the claim that the ECA ‘refuses to sign off the Commission’s accounts’ is incorrect and misleading.  As this paper states the problems are real ones – as they are in the Member States’, much larger, national spending programmes.  But they are being addressed by the EU’s institutions.

  1.   European Court of Auditors, Annual Report on the Implementation of the Budget 2013, 2014 OJ C 398/01, p. 13 (hereinafter ECAAR 2013)
  2.   Consolidated Version of the Treaty on the Functioning of the European Union, art. 317, 2012 OJ C 326/47 at p. 186 (hereinafter TFEU)
  3.   Ibid., art. 287(1), p. 170
  4.   ECAAR 2013, p. 11
  5.   See, e.g., ECAAR 2013, pp. 17-19
  6.   ECAAR 2013, p. 11
  7.   Ibid., p. 12
  8.   This is a simplified description of the ECA’s method; a fuller account can be found in ECAAR 2013, Annex 1.1, p. 42, et seq.
  9.   The same is true if one looks not just at MLE but at ‘95% confidence intervals’, the range within which the Court is 95 per cent confident that the true rate of error in payments lies. For the budget as a whole in 2013 the Court was 95 per cent confident that quantified error lay between 3.5 per cent and 5.9 per cent. See ECAAR 2013, Table 1.2, p. 22 for the 95 per cent confidence intervals for individual broad areas of expenditure.
  10.   European Court of Auditors, Agriculture and cohesion: overview of EU spending 2007-2013, 9 December 2014 (hereinafter Overview Report)
  11.   Overview Report, p. 17
  12.   ECAAR 2013, p. 22
  13.   ECAAR 2013, p. 25
  14.   2007 and 2008 omitted because payment then related mainly to an earlier period in the EU’s multiannual financial plans.
  15.   Overview Report, pp. 17, 43
  16.   ECAAR 2013, p. 45
  17.   Overview Report, p. 43
  18.   Overview Report, pp. 22, 43
  19.   European Commission, Synthesis of the Commission’s management achievements in 2014, COM (2015) 279 final, 3 June 2015, p. 14
  20.   The Overview Report (p. 27) states that 2007-2013 agriculture and cohesion expenditure ‘financial corrections’ imposed by the Commission did not in practice reduce payments to final beneficiaries.  For example, for cohesion, national authorities were entitled to declare additional projects which compensated for the finan­cial corrections applied.  The Commission then, after the normal checks, paid the Member State concerned the difference between the new claim(s) and the ‘financial correction’.  On the other hand, the Commission’s ‘disallowance’ of expenditure on agriculture has the effect of shifting the cost of the expenditure concerned from the EU budget to the Member State concerned.
  21.   National Audit Office, Department for Work and Pensions: 2012-13 Accounts: Report by the Comptroller and Auditor General, 9 December 2013
  22.   US Government, Financial Report Fiscal Year 2014 of the United States Government, 26 February 2015, p. 25
  23.   TFEU, art. 235, p. 152
  24.   European Commission, Protection of the European Union’s financial interests – Fight against fraud 2013: Annual Report, COM (2014) 474 final, 17 July 2014, p. 13
  25.   European Anti-Fraud Office, The OLAF Report 2014: Fifteenth report of the European Anti-Fraud Office, 26 May 2015, p. 9
  26.   European Anti-Fraud Office, supra n. 25, pp. 11-24
  27.   Ibid., pp. 25-31