The Commission’s response to fraud in the Common Agricultural Policy – Time to dig deeper

Executive summary

I Common Agricultural Policy (CAP) spending supports agriculture and rural development in the EU through:

• direct payments to farmers, in general based on the area of agricultural land at the beneficiaries’ disposal;
• agricultural market measures, helping EU agricultural sectors to adapt to market changes;
• Member States’ national and regional rural development programmes supporting social and economic development in rural areas, and providing aid based on environmental and climate-related criteria.

II In 2018-2020, in the EU-27 direct payments amounted to an average of €38.5 billion each year; market measures and rural development spending averaged €2.7 and €13.1 billion respectively.

III Fraud harms the EU’s financial interests and prevents EU resources from achieving the policy objectives. We expect our report to help the Commission and the Member States to develop their anti-fraud capacity under the new common agricultural policy 2023-2027.

IV Various bodies protect CAP financing from fraud at EU and national level. Our audit examined whether the Commission has taken appropriate action on fraud in CAP spending, by assessing the fraud risks inherent in CAP payment schemes, and whether the Commission has identified and properly responded to the fraud risks affecting CAP spending.

V We found that the Commission has responded to instances of fraud in CAP spending, but was not sufficiently proactive in addressing the impact of the risk of illegal land grabbing on CAP payments, in monitoring Member States’ anti-fraud measures, and in exploiting the potential of new technologies.

VI Fraud risks vary between the CAP payment schemes. In this report, we present an overview of the fraud risks affecting the various CAP payments schemes. We identified risks linked to beneficiaries concealing breaches of eligibility conditions, to the complexity of the financed measures, and to illegal forms of ‘land grabbing’.

VII The Commission assessed the fraud risks in CAP spending, and recognised rural development investment measures and certain market measures as presenting a higher risk than other payment schemes.It has recently issued some guidance to paying agencies related to ‘land grabbing’ and the notion of ‘land at the farmer’s disposal’.

VIII The Commission provided Member States with guidance on fraud-related issues. The majority of paying agencies appreciated this guidance, but some considered that it could usefully have included more practical examples. The latest CAP fraud risk analysis by the Commission dates back to 2016, and the Commission plans to carry out a new one before the new CAP enters into force in January 2023.

IX The Commission carries out accreditation and compliance checks of paying agencies to assess their control systems, which may cover anti-fraud measures. The Commission relies on the certification bodies’ annual reviews to monitor paying agencies’ compliance with their accreditation criteria, including anti-fraud measures. Some certification bodies’ reports provided little analysis of the paying agencies’ anti-fraud measures, but the Commission did not require the certification bodies to provide further details in these cases.

X The Commission has promoted the use of new technologies to automate checks, as in the case of ‘checks-by-monitoring’ that monitor the whole population of aid recipients for a given scheme, and has developed its own risk-scoring tool, Arachne, to support Member States in preventing fraud. The use of these technologies is voluntary, and Member States have responded at a low pace. Artificial intelligence and big data have potential in fighting fraud but Member States face challenges in seizing these opportunities and the Commission has started promoting these technologies.

XI We make recommendations to the Commission so that it can gain and share a deeper insight of fraud risks and measures in CAP spending to protect EU financial interests, and promote the use of new technologies in preventing and detecting fraud.

Introduction

The Common Agricultural Policy (CAP)

01 CAP spending supports agriculture and rural development in the EU through:

• direct payments to farmers, fully funded by the EU budget, and in general based on the area of agricultural land at the beneficiaries’ disposal;
• agricultural market measures, also fully funded by the EU budget with the exception of certain measures co-financed by the Member States; market measures include support for public and private storage of agricultural products in case of market disturbances, sector-specific aid schemes (e.g. in the wine or fruit and vegetables sectors), and the reimbursement of the costs of promoting sales of EU agricultural products;
• Member States’ national and regional rural development programmes, co-financed by the EU budget and the Member States, involving both reimbursement of project costs, and payments based on beneficiaries’ area of agricultural land or number of animals.

02 In 2018-2020, in the EU-27 direct payments amounted to an average of €38.5 billion each year; market measures and rural development spending averaged €2.7 and €13.1 billion respectively.

03 The EU farming sector shows significant variations in terms of the types of beneficiary, the size of holdings, and the forms of tenure of agricultural land. This makes it challenging to design rules and control systems, which suit all scenarios.

04 CAP beneficiaries range from private individuals to cooperatives, companies and public sector bodies. In the 2020 financial year around 6.2 million beneficiaries received direct payments, 3.5 million (typically also recipients of direct payments) received payments under rural development measures, and 102 000 received market measure support.

05 Most CAP beneficiaries receive less than €10 000 per year: the proportion has been decreasing, but still represents more than 80 % of beneficiaries (see Figure 1).

06 Concerning the structure of agricultural holdings, 67 % of EU farms are below 5 hectares, while 3 % have more than 100 hectares. The situation varies among Member States. For example, in Malta and Romania, more than 90 % of farms are below 5 hectares, whereas in Denmark and Finland only 4 % of farms fall into this category.

07 The distribution of agricultural land also varies. In Slovakia, 9 % of farms are larger than 100 hectares and cover 89 % of the national agricultural area, whereas in Slovenia such farms cover only 7 % of the agricultural land. In Figure 2 we show the situation at EU level, in the three selected Member States (see paragraph 26) and in the two Member States at the extreme of the spectrum.

08 As regards the tenure of agricultural land, 82 % of Polish farmers own their land, whereas 78 % of Maltese farmers rent it, and in Greece one third of the agricultural area is common land (see Annex I).

Protecting the CAP budget against fraud

09 EU legislation¹ defines fraud affecting the EU’s financial interests as a deliberate infringement (act or omission) that is, or could be, prejudicial to the EU budget through:

• the use or presentation of false, incorrect or incomplete statements or documents,
• non-disclosure of information in violation of a specific obligation, or
• the misapplication of funds for purposes other than those for which they were originally granted.

10 The key distinguishing factor between fraud and irregularity is the notion of intent. An irregularity may be the result of an incorrect interpretation of a rule, whereas fraud is the result of a deliberate breach of a rule.

11 Fraud can be categorised as ‘internal’ or ‘external’². In the CAP context:

• internal fraud may be committed by staff of public authorities involved in the administration of CAP funds or by staff of EU institutions or bodies. It may involve undeclared conflicts of interest, breaches of professional confidentiality, or passive corruption;
• external fraud refers to fraud committed by beneficiaries of CAP funding. Examples are public procurement fraud (e.g. collusion between tenderers, irregular or fictitious subcontracting, active corruption), falsification of documents, inflation of costs, or hiding links between companies.

12 Fraud harms the EU’s financial interests and prevents EU resources from achieving the policy objectives. Fighting fraud effectively requires a comprehensive risk management framework, covering the full anti-fraud cycle of prevention, detection and response to fraud (see Figure 3).

Various bodies protect CAP financing from fraud

13 Within the Commission:

• the Directorate-General for Agriculture and Rural Development (DG AGRI) shares the CAP’s management with accredited paying agencies in the Member States, while remaining ultimately responsible for the policy. DG AGRI obtains assurance on the operation of the Member States’ management and control systems;
• the European Anti-fraud Office (OLAF) carries out administrative investigations into illegal activities adversely affecting the EU budget, and serious misconduct within the EU institutions. OLAF hosts the Irregularities Management System (IMS), containing data submitted by Member States on irregularities and fraud in EU spending. OLAF analyses the data provided by the Member States and publishes the Commission’s ‘Annual Report on the Protection of the European Union’s financial interests - Fight against fraud’ (‘PIF Report’)³. OLAF plays a central role in the development of the EU anti-fraud policy.

14 In June 2021, the European Public Prosecution Office (EPPO) began operations as the EU’s independent public prosecutor. EPPO has the power to investigate, prosecute and bring to judgement crimes against the EU budget, such as fraud and corruption, in the competent courts of the 22 Member States participating in EPPO⁴.

15 Member States are required to take measures to prevent, detect and correct fraud and other irregularities, and protect the EU’s financial interests in the same way as their own financial interests. They must report cases of irregularities and fraud exceeding €10 000 to the Commission through the IMS.

16 As part of their accreditation criteria, paying agencies are required to set up internal control activities designed to prevent and detect fraud (see Box 1). Directors of the paying agencies provide the Commission with an annual management declaration on the proper functioning of the internal control systems, where they have to confirm that effective and proportionate anti-fraud measures are in place, which take account of the risks identified⁵.

17 While paying agencies are not responsible for investigating fraud, they should take measures to prevent and detect fraud and cooperate with enforcement authorities undertaking investigations.

18 Certification bodies designated by Member States review yearly the paying agencies’ internal control systems, and their compliance with the accreditation criteria. They provide an annual opinion on the legality and regularity of the expenditure.

19 Figure 4 shows the roles and responsibilities of the main bodies combating fraud in the CAP.

20 In line with auditing standards, we are also vigilant to fraud risks. In particular:

• we take account of the risk of fraud when planning and performing our audits;
• we report cases of suspected fraud to OLAF and EPPO;
• we have published two recent special reports on combatting fraud in EU spending⁶, and we examined the Commission’s anti-fraud policies and procedures in relation to CAP spending in our 2019 Statement of Assurance work⁷.

Member States report to the Commission the amount of fraud they detect in CAP spending

21 The Commission’s PIF report provides an overview of fraud in CAP spending, in terms of number of reported cases and financial amounts, and complements the statistics with underlying analyses. From 2016 to 2020, the CAP accounted for 11 % of fraudulent amounts reported in IMS, whereas Cohesion policy represented 86 % of the total (see Table 1).

22 The financial impact of the reported fraudulent irregularities for the CAP is generally low: in 2016-2020, it amounted to 0.09 % of total CAP payments. Figure 5 shows the fraud detection rate, i.e. the ratio between the reported amounts of fraud cases (suspected and established) and the payments for the CAP components.

23 However, as explained in the 2020 PIF report⁸, the number of irregularities reported as fraudulent – and the associated amounts - are not a direct indicator of the level of fraud affecting the EU budget, but rather of the Member States’ work to counter fraud and other illegal activities affecting the EU’s financial interests. In our previous audits, we found that these figures did not provide a complete picture of the detected fraud level in EU spending⁹.

Audit scope and approach

24 Our audit examined patterns of fraud in CAP payment schemes. We assessed whether the Commission has properly identified the fraud risks affecting CAP spending, and whether it has adequately responded to those risks.

25 We covered the measures financed by the CAP under shared management (direct payments, market measures and rural development), examining data from the 2007-2013 and 2014-2020 programing periods. We did not examine CAP spending under direct management (around 0.8 % of CAP spending), and we did not cover conflict of interest, which is the subject of a separate ECA special report to be published in 2022. The EPPO, which became operational in June 2021, was not included in the scope of our audit.

26 We obtained our evidence from:

• documentary reviews and video-conferences at OLAF and at the Directorate-General for Agriculture and Rural Development;
• documentary reviews in three Member States (France, Italy and Slovakia). Their selection was based on indicators relating to fraud, land concentration, and funding amounts;
• surveys addressed to the Supreme Audit Institutions of the three selected Member States, and to the paying agencies, anti-fraud coordination services (AFCOS) and certification bodies in all 27 Member States. The survey covered fraud risks, anti-fraud measures and controls in place in the Member States. AFCOS and paying agencies from 23 Member States and certification bodies from 13 Member States replied to our survey;
• analysis of data extracted from the IMS and from datasets of the EU statistical office (Eurostat).

27 We expect our report to help the Commission and the Member States to develop their anti-fraud capacity under the new common agricultural policy 2023-2027.

Observations

Fraud risks vary between the CAP payment schemes

28 When proposing legislative acts, the Commission is responsible for ensuring that the design and rules of the various CAP support schemes take into account the inherent risk of fraud¹⁰.

29 We assessed the fraud risk exposure of the main categories of CAP spending. We took into account the results of our previous audits, together with cases reported in IMS, and OLAF investigations.

30 Our Statement of Assurance work has shown that the complexity of rules and the way EU funds are disbursed have an impact on the risk of error. In our 2019 annual report¹¹, we observed that the risk of fraud is also greater in spending areas subject to more complex eligibility conditions (see Figure 6).

Some beneficiaries conceal breaches of eligibility conditions

31 Some CAP payment schemes aimed at supporting specific categories of beneficiaries have proven susceptible to fraud, as some claimants fail to disclose relevant information or artificially create the conditions to meet the eligibility criteria and unduly benefit from CAP aid.

32 One of the EU priorities for rural development is facilitating the diversification, creation and development of small and medium-sized enterprises (SMEs). In order to establish whether a beneficiary is eligible as an SME, the disclosure of reliable information on the number of employees, annual turnover and annual balance sheet total is essential, as is the disclosure of information on linked companies.

33 In our statement of assurance work, we have identified cases of beneficiaries who failed to disclose their links with other companies. OLAF has also investigated such cases (see examples in Box 2).

34 The CAP seeks to encourage generational renewal in agriculture by granting additional funding to persons qualifying as ‘young farmers’. In order to be eligible for such support, farmers must be no more than 40 years old when submitting their aid application and must set up for the first time an agricultural holding as head of the holding.

35 A young farmer may set up a holding solely or jointly with other farmers, irrespective of the legal form. However, in the case of legal persons, a young farmer must exercise effective and long-term control over the entity in terms of decisions related to management, benefits and financial risks.

36 Member States can define stricter rules for young farmers receiving rural development start-up support. For instance, in France, national eligibility rules require that the young farmer’s income must come mainly from agricultural activities. In IMS, some cases showed that when the new farm did not generate the expected turnover, young farmers found other jobs (sometimes full-time) to increase their income, thus becoming ineligible. Other cases showed that beneficiaries were not the head of the agricultural holding, nor actually working on the farm.

37 Such cases may be considered fraud when a beneficiary, who does not meet the eligibility conditions, presents false or incomplete information to misrepresent their true situation.

38 In some cases, beneficiaries may falsify documents or simulate activities in order to qualify for CAP aid (see Box 3).

Fraud risks increased by the complexity of the projects

39 Within market measures, the wine support programmes finance a number of measures, subject to a variety of eligibility conditions, such as restructuring and conversion of vineyards, harvest insurance, investment in enterprises, innovation for the development of new products, processes and technologies, and promotion in non-EU countries.

40 Promotional actions in non-EU countries are especially risk-prone. These activities can be challenging to check, due to:

• the transient and intangible nature of many promotional activities (e.g. public relations);
• on-the-spot checks in non-EU countries being rare;
• most promotional actions being sub-contracted, especially when taking place in countries other than that of the beneficiary.

41 Box 4 shows an example detected during our Statement of Assurance work.

Direct payments and ‘land grabbing’

42 In order to receive area-based direct payments, beneficiaries must declare the eligible number of hectares ‘at their disposal’¹⁸. This means that beneficiaries must have a proper legal basis for claiming the land.

43 Direct payments have been associated with the term ‘land grabbing’, although the term is controversial (see Box 5).

44 In replying to our survey, almost 60 % of paying agencies indicated that they did not consider ‘land grabbing’ as a risk indicator. Five paying agencies in one Member State associated ‘land grabbing’ with the situation where eligible claims were submitted by beneficiaries without having a legal basis to claim the land.

Current controls have contributed to decreasing the risk of error

45 The main management tool to check the eligibility of direct payments is the Integrated Administration and Control System (IACS), incorporating the Land Parcel Identification System (LPIS).

46 IACS interlinks databases of holdings, aid claims, agricultural areas and animal registries, which the paying agencies use to perform administrative cross-checks on all aid applications. LPIS is a geographical information system containing multiple-source spatial datasets, which together form a record of agricultural areas in the Member States.

47 As confirmed by our Statement of Assurance work, IACS, and LPIS in particular, form an effective management and control system to ensure that direct aid payments as a whole are not affected by material error.

48 The introduction in IACS of the geo-spatial aid application (GSAA), which enables farmers to submit payment claims online, and the fact that paying agencies carry out preliminary cross-checks on farmers’ aid applications, have also contributed to reducing the level of error.

Some situations are more prone to ‘land grabbing’

49 Since IACS-LPIS makes it difficult to overstate the eligible area (e.g. via double claims or claiming non-agricultural land), fraudsters seek to acquire agricultural land illicitly and then claim support (see Figure 7).

50 Investigations by OLAF and national authorities have found that the agricultural areas most susceptible to this type of fraudulent activity are publicly owned land or private land with unclear ownership (see Box 6 and Box 7).

51 Fraudsters illicitly claiming land for direct payment support may submit forged documents, and resort to criminal practices, such as extortion and collusion with officials (internal fraud). Box 6 shows an example.

52 Fraudsters may also exploit weaknesses in Member States’ checks (see Box 7).

The Court of Justice and the Commission have recently clarified the rules applicable to checks on the legal basis to use the land

53 CAP legislation does not define the concept of ‘land at the farmer’s disposal’, nor require farmers to provide proof of their right to use the land when submitting an aid application²⁴. National rules regarding ownership, leases or other forms of legal tenure apply.

54 Member States must perform checks on all claims to prevent and correct irregularities and recover undue payments²⁵, and to this end, they may require the claimant to provide evidence that the land is legally at their disposal²⁶, in particular in case of doubt.

55 Checks on farmers’ legal right to use the claimed land vary between Member States. According to the results of our survey, eight paying agencies in two Member States check this in all cases. Nine out of 47 paying agencies replying to our survey declared that they make checks only in case of overlapping claims, whereas the remaining paying agencies also carry out targeted checks in other situations: e.g.

•  in case of land claimed for the first time or owned by a public entity,
•  in case of doubts raised during administrative checks or field visits.

56 When land is claimed by more than one person purporting to have a legal basis to make the claim, the principle of ‘decision-making power’, in terms of benefits and financial risks borne by the farmer, applies²⁷. A recent ruling of the European Court of Justice has introduced some clarifications on the issues at stake, drawing attention to the relevance of the land being lawfully at the applicant’s disposal. The Court ruled that when land is claimed both by the owner and by a third party using the land without legal basis, the land is considered at the disposal of the owner²⁸.

57 Following this judgement, in June 2021, the Commission issued guidance to the Member States²⁹. The guidance explains that for a beneficiary to have land lawfully at their disposal implies obtaining the right to use it in a lawful way. The guidance also states that Member States can design their own checks, but that those checks must prevent and correct irregularities effectively, and should not be limited to double claims.

Funds claimed for land without performing any agricultural activity

58 Fraudsters may also seek to acquire land – legally or not – for the sole purpose of receiving direct payments, without performing any agricultural activities. The risk is higher for certain pastureland and mountainous areas, where it is more difficult for paying agencies to check the required agricultural activity, such as grazing (see Box 8).

Our reporting of suspected fraud cases

59 During our Statement of Assurance work each year, we detect a number of potentially fraudulent irregularities, but we cannot be sure that fraud had actually taken place.

60 In 2018-2020, the overall error rate we reported for spending on ‘Natural resources’ decreased from 2.4 % to 2.0 % of the total paid. The CAP accounted for around 97 % of spending on ‘Natural resources’. Over those years we audited 698 CAP payments and quantified errors in 101 cases. In 17 of those cases, we suspected that the error could be associated with fraud.

61 We are not entitled to investigate fraud, and we forwarded 12 of the 17 cases to OLAF, while in the remaining five cases investigations or recovery procedures were already ongoing, or the low amount involved would not meet OLAF’s criteria for investigation on grounds of proportionality.

The Commission has taken action on fraud in CAP spending but was not sufficiently proactive

62 In order to fight fraud, we would expect the Commission to take appropriate measures to obtain an overview of fraud and fraud risks in CAP spending, and properly respond to them³¹.

63 Under its Anti-Fraud Strategy³², DG AGRI has taken action to reinforce its cooperation with OLAF (in terms of reporting cases to OLAF and following up OLAF recommendations), deliver training and guidance to its staff, provide guidance to Member States and raise fraud awareness internally and externally.

64 We examined whether the Commission has:

• carried out a comprehensive fraud risk assessment, and monitored the emergence of new fraud typologies;
• provided appropriate guidance and raised Member States’ awareness on identified fraud risks;
• adequately monitored Member States’ anti-fraud measures;
• promoted the use of new technologies to reinforce control systems.

The Commission has identified key fraud risks and recently issued some guidance related to ‘land grabbing’

65 One of the strategic objectives set out in DG AGRI’s Anti-Fraud Strategy was ‘reinforcing fraud risk assessment’. DG AGRI reported on its analyses in successive versions of its Anti-Fraud Strategy (see Table 2). DG AGRI considered:

• rural development investment measures and certain market measures (promotion actions and support to producer organisations) as higher risk,
• direct payments, other area-related payments and animal-related payments as less risk-prone.

66 From 2014, DG AGRI undertook to monitor emerging fraud typologies. In 2016, it carried out an extensive fraud risk analysis and committed to update it annually. Since 2017, DG AGRI has considered that an update of its risk analysis has not been needed, on the basis that the risks had not changed and no new fraud patterns had emerged. DG AGRI plans to carry out a new fraud risk analysis before the new CAP enters into force in January 2023.

67 In 2017-2019, OLAF investigated cases of systematic abuses regarding direct payments in Italy, France and Slovakia (see Box 6, Box 7 and Box 8). DG AGRI did not complement its fraud risk assessment to cover ‘land grabbing’, as it did not consider it as direct fraud against the CAP.

68 In its 2020 Anti-Fraud Strategy, DG AGRI described ‘land grabbing’ as the illicit appropriation of agricultural land for which fraudsters then claim direct payments in a legal manner. DG AGRI stated that the strategy also covered reputational risks without direct impact on the EU’s financial interests, which may include ‘land grabbing’ and other misbehaviours. In this context, DG AGRI assessed that these phenomena do not originate from weaknesses of CAP legislation, but are a matter of rule of law in the Member States concerned.

69 As we reported in Box 5, ‘land grabbing’ can be linked to a range of fraudulent practices such as the falsification of documents, coercion, use of political influence or insider information, manipulation of procedures or payment of bribes. According to the Commission’s guidance based on recent case law (see paragraph 57), for a beneficiary to have land lawfully at their disposal implies obtaining the right to use it in a lawful way. As a consequence, when it is established, in accordance with national law, that such a right has been obtained by resorting to fraudulent practices, this renders the underlying payments irregular. In such cases, paying agencies can play a role by using the data at their disposal to identify red flags and cooperating with law enforcement authorities undertaking investigations.

The Commission has provided guidance to Member States, but some paying agencies indicated a need for more practical advice

70 DG AGRI took actions aimed at raising the Member States’ awareness on fraud-related issues and provided them with guidance and fraud indicators (‘red flags’). Figure 8 summarises these activities.

71 Replying to our questions on the guidance received from the Commission, paying agencies mentioned DG AGRI’s initiatives, but also referred to other sources of information:

• the Commission’s anti-fraud guidance for the European Structural and Investment Funds;
• information exchanged within the Advisory Committee for the Coordination of Fraud Prevention (COCOLAF) and within the DG AGRI Learning Network;
• ‘PIF Reports’ and other guidance provided by OLAF.

72 The majority of the surveyed paying agencies found the Commission’s guidance helpful, although two of them remarked that some analyses (e.g. in the PIF reports) are interesting but too generic to be useful in their daily work. Seven paying agencies appreciated the examples of cases of fraud, which were provided during the anti-fraud seminars. Two paying agencies considered that the guidance should be complemented by more practical examples.

73 Figure 9 shows the guidance provided by DG AGRI since 2012.

74 DG AGRI did not issue any guidance notes on fraud risk indicators (‘red flags’) after 2013, although the establishment of additional fraud indicators would be beneficial in guiding paying agencies in their work. Most dedicated anti-fraud seminars took place between 2013 and 2018; they started again in 2021. From 2020, Member States and DG AGRI regularly discussed anti-fraud measures during their Learning Network meetings.

The Commission monitors the anti-fraud measures implemented by the Member States, but its overview has gaps

75 In order to conform to the fraud-related accreditation criteria set in EU legislation³³ (see Box 1), in 2014 the Commission (DG AGRI) recommended paying agencies adopt specific measures (see Box 9).

76 Every year, certification bodies in the Member States review paying agencies’ compliance with the accreditation criteria and their internal control systems. According to the Commission, they should also assess Member States’ implementation of the recommended anti-fraud measures (see Box 9). The Commission examines the annual reports of the certification bodies, and follows up identified weaknesses.

77 In 2016, the Commission provided certification bodies with guidelines on the accreditation criteria, including a section on fraud prevention and detection, but did not give any guidance on the checks that certification bodies were expected to carry out.

78 We reviewed certification body reports from 2018 to 2020 on five paying agencies. Three certification bodies reported on their checks and made recommendations when they found weaknesses, whereas the remaining two certification bodies did not include such information in their reports. The Commission has not required these two certification bodies to provide any further details.

79 Incomplete review and reporting by certification bodies may impair the assurance on the quality of the anti-fraud measures in place in the paying agencies. In a meeting with certification bodies which took place in November 2021 the Commission further discussed their role in this respect.

80 DG AGRI carries out compliance and accreditation checks to assess paying agencies’ control systems, which may cover anti-fraud measures. With regard to accreditation checks, we reviewed DG AGRI accreditation reports for the Member States we selected and found that:

• for France, no recent accreditation checks were carried out for the three selected paying agencies,
• for Italy, enquiries were carried out in 2015 and 2019 in the selected paying agency, but they focused on debt management rather than anti-fraud measures,
• for Slovakia, DG AGRI focused on anti-fraud measures (see next paragraph and Box 10).

81 In specific cases, where the Commission becomes aware of possible misuse of funds, the paying agency may be subject to in-depth checks (see Figure 10 and Box 10).

The Commission has promoted the use of new technologies but these are not sufficiently exploited

82 The Commission has promoted the use of new technologies in carrying out administrative checks through IACS. It has also encouraged the use of ‘checks-by-monitoring’ and Arachne.

83 Since 2018, paying agencies may perform ‘checks-by-monitoring’. This approach uses automated processes to check compliance with CAP rules for features that can be monitored based on satellite data. So far, paying agencies have mainly used ‘checks-by-monitoring’ to assess area-based aid claims under direct payment schemes.

84 The paying agencies can compare satellite data on crop types and agricultural activity with information provided by farmers in their aid applications. Where all eligibility criteria of a given payment scheme can be evaluated from space, it enables paying agencies to monitor the whole population of claimed parcels remotely.

85 The new approach enables paying agencies to warn farmers of potential non-compliance during the growing season (for example, to mow a field by a certain date). This provides farmers with more opportunities to correct their claims before they are finalised, and encourages compliance with the scheme rules³⁵.

86 ‘Checks-by-monitoring’ have the potential to reduce administrative burden and improve cost-effectiveness³⁶. Providing regular observations of the agricultural activity of an entire population of beneficiaries, they have a deterrent effect and can help identify red flags for potential fraud.

87 Since 2013, the Commission has developed its own risk-scoring IT tool, Arachne, available to the Member States free of charge and on a voluntary basis. Arachne processes and analyses data provided by Member States on beneficiaries, contractors and other stakeholders related to a project, and cross-checks the data with information from external databases on companies and persons linked to those companies. This can identify projects, beneficiaries and contractors at risk of fraud.

88 Originally developed for the cohesion spending area, Arachne has been extended to the CAP through a pilot project launched in February 2019 covering rural development projects.

89 The legislation governing the CAP from 2023³⁷ requires the Commission to make the tool available for voluntary use by Member States. The Commission is required to publish an assessment report in 2025. Through a joint statement, the Council and the European Parliament committed to examining a proposal on the compulsory use of the tool, following the Commission’s assessment report³⁸.

Member States’ implementation of these technologies takes time

90 In our special report on the use of new imaging technologies to monitor the CAP, we recommended the Commission to promote ‘checks-by-monitoring’ as a key control system for the post-2020 CAP³⁹, and the Commission has committed itself to providing support to the Member States in developing this new approach.

91 At the end of 2020, two years from the start-up of the tool, ‘checks-by-monitoring’ covered 5.7 % of the total area receiving direct payments⁴⁰. According to Commission estimates, by the end of 2021 this coverage had reached 13.1 %.

92 Under the new CAP, automated analysis via satellite data will be mandatory for area-based measures in all Member States through the ‘Area Monitoring System’ (AMS). The Commission expects this requirement to increase the area under satellite monitoring.

93 Two years after the launch of the Arachne pilot project for the CAP, seven⁴¹ out of 76 paying agencies have uploaded data and use the tool to some extent, four⁴² are in testing mode and ten⁴³ have initiated preliminary discussions with a view to its use. Since Arachne is a risk-scoring tool based on data-mining, its usefulness depends on how much data Member States’ authorities upload and on whether it is used.

Potential for further technological development needs to be exploited

94 Artificial Intelligence has significant potential to improve working tools, enabling the detection of patterns among billions of data points. Data-mining tools may make monitoring systems more efficient and able to detect fraud and mismanagement of public funds.

Use of big data in identifying beneficial owners

95 Public disclosure and reporting requirements are essential to allow for accountability and scrutiny to prevent corruption and fraud.

96 In May 2021, the European Parliament published a study⁴⁴, analysing the beneficiaries receiving CAP funds in 2018 and 2019 and cohesion funds between 2014 and 2020, in order to identify the largest beneficiaries. The study distinguished between ‘direct beneficiaries’ – the direct recipients of EU funds – and ‘ultimate beneficiaries’ (beneficial owners), i.e. legal or natural persons directly or indirectly controlling the largest share of the direct beneficiary. The study highlighted the technical and legal difficulties of obtaining a comprehensive overview of the ultimate beneficiaries and the amounts of EU funds received.

97 The report showed that public bodies, limited liability companies and other legal persons made up about one-tenth of the direct beneficiaries, but received more than one-third of CAP funds in 2018-2019. To overcome the fragmentation and challenges in identifying the ultimate beneficiaries of EU funds, the report suggested the creation of a common EU database including all projects financed by the CAP and the European Structural and Investment Funds. In September 2021, the European Parliament published a study on the requirements for a unique beneficiary database⁴⁵.

98 A common database of EU beneficiaries would gather data on millions of operators. These big data would allow the identification of patterns in the distribution of funds and contribute to alert on potential fraudulent situations.

99 As stressed by the Commission in its 2020 PIF report⁴⁶, improved transparency about beneficiaries (including contractors, sub-contractors and beneficial owners) of public (EU and national) funding, and a more efficient collection and use of data, fully exploiting the opportunities offered by IT interconnectivity, data mining and risk-scoring tools, are pivotal to the fight against fraud.

100 The 2021-2027 common provisions regulation in the field of cohesion⁴⁷ requires Member States to collect information on the beneficial owners of the recipients of EU funding. The new CAP legislation requires Member States to collect the information necessary to identify beneficiaries, including the identification of the group in which they participate, but it does not mention information on beneficial owners⁴⁸.

Use of artificial intelligence in identifying fraud risks

101 The Commission recommends paying agencies to set up a ledger of fraud risk indicators (‘red flags’). Replying to our survey, paying agencies provided some examples (Figure 11).

102 Some red flags, especially those concerning direct payments, could be incorporated directly into the IACS database. For instance, the Italian authorities plan to upload in their system details of parcels, which have been confiscated, or are subject to judicial proceedings, so that the system may raise an alert.

103 Machine-learning techniques could be exploited to further automate checks and raise red flags. The paying agency in Estonia has applied machine-learning techniques to satellite images to prevent and detect non-compliances with mowing requirements. The paying agency considers that this system has increased awareness among beneficiaries that its monitoring activity covers 100 % of the parcels, leading to a deterrent effect, and contributing to the prevention of irregularities and fraud. Another paying agency has implemented a system based on an algorithm which looks for indicators of parcels at risk of irregular claims. The paying agency then identifies companies with a significant number of such parcels, where there may be an increased risk of irregularity and fraud.

104 In Italy, a joint initiative between the national paying agency and the Interior Ministry aims to set up a profiling model to identify municipalities with a high probability of criminal focus areas, by using data mining and photointerpretation. Examining satellite images available in the LPIS, the project analyses qualitative and quantitative variations of the territorial elements potentially attributable to criminal activities, such as illegal landfills, spills, slums, illegal artefacts, and abandoned buildings. Comparing this information with data on the municipalities, the projects aims to identify situations where criminal activities have occurred or may potentially occur.

105 Artificial intelligence has great potential but requires high volumes of data, in accessible format, and interconnection of databases. Replying to our survey, some paying agencies highlighted obstacles to their access to data, such as:

• the exchange of information between different actors and databases may be hindered by legal requirements on banking secrecy or data protection;
• the interconnection of databases and automation of processes and controls require specific IT capacity and expertise;
• developing the necessary IT capabilities requires substantial financial investment, which may be disproportionate in relation to the amount of funds allocated to some measures, small Member States or the level of potential fraud.

Conclusions and recommendations

106 Our audit examined whether the Commission has taken appropriate action on fraud in Common Agricultural Policy (CAP) spending. We assessed patterns of fraud in CAP payment schemes. We examined whether the Commission has identified and properly responded to the fraud risks affecting CAP spending.

107 Our overall conclusion is that the Commission has responded to instances of fraud in CAP spending, but was not sufficiently proactive in addressing the impact of the risk of illegal land grabbing on CAP payments, in monitoring Member States’ anti-fraud measures, and in exploiting the potential of new technologies.

108 In this report, we presented an overview of the fraud risks affecting the CAP. We identified risks linked to beneficiaries concealing breaches of eligibility conditions (paragraphs 31 to 38), to the complexity of the financed measures (paragraphs 39 to 41), and to illegal forms of ‘land grabbing’ (paragraphs 42 to 58).

109 The Commission assessed the fraud risks in CAP spending, and recognised rural development investment measures and certain market measures as most at risk (paragraph 65). The Commission carried out its latest CAP fraud risk analysis in 2016, and plans a new analysis before the new CAP enters into force in January 2023 (paragraph 66).

110 The Commission provided Member States with guidance on fraud-related issues (paragraph 70), and the majority of paying agencies appreciated this guidance (paragraph 72). From 2020, the Commission discussed anti-fraud measures with Member States during regular meetings (paragraph 74), and in 2021, it issued specific guidance on the checks paying agencies need to carry out on the legal basis of the claimed agricultural land (paragraph 57).

111 The Commission’s accreditation and compliance checks of paying agencies may cover anti-fraud measures (paragraph 80). The Commission relies on the certification bodies’ annual reviews to monitor paying agencies’ compliance with their accreditation criteria, including anti-fraud measures (paragraph 76). In 2016, the Commission provided certification bodies with guidelines on the accreditation criteria, but did not give any guidance on the checks that certification bodies were expected to carry out in relation to fraud (paragraph 77). Some certification bodies’ reports provided little analysis of the paying agencies’ anti-fraud measures, but the Commission did not require the certification bodies to provide further details in these cases (paragraph 78).

112 The Commission has promoted the use of new technologies to automate checks, as in the case of ‘checks-by-monitoring’ based on satellite images (paragraphs 83 to 86), and has developed its own risk-scoring tool, Arachne, to support Member States in preventing fraud (paragraphs 87 to 89). Member States have taken up these technologies at a low pace (paragraphs 91 and 93). Artificial intelligence and big data have potential in fighting fraud (paragraphs 94 to 104) but Member States face challenges in seizing these opportunities (paragraph 105) and the Commission has started promoting these technologies.

Recommendation 1 – Gain and share a deeper insight of fraud risks and measures in CAP spending

The Commission should:

1. update the guidelines for certification bodies to clarify their role in assessing paying agencies’ anti-fraud measures, and check how certifications bodies follow the guidelines;
2. review how paying agencies implement the guidance on checking that land is at the applicants’ lawful disposal, and disseminate best practices addressing the risks of illegal land grabbing;
3. on the basis of a) and b), update its assessment of the extent to which different spending schemes are exposed to fraud risks and the extent to which anti-fraud measures implemented at Member State level are able to detect, prevent and correct them, and take the necessary measures to mitigate key fraud risks.

Target implementation date: 2023

Recommendation 2 – Promote the use of new technologies in preventing and detecting fraud in CAP spending

The Commission should:

1. support paying agencies in their use of ‘checks-by monitoring’ and the future ‘area monitoring system’ by identifying bottlenecks in the uptake of new technologies and sharing best practices and technical solutions on how to remove these bottlenecks;
2. promote the use of fraud-detection tools, such as Arachne, among paying agencies, in order to increase the number of Member States using them;
3. share with Member States best practices on the use of artificial intelligence and machine-learning to identify patterns indicating fraud.

Target implementation date: 2024

This Report was adopted by Chamber I, headed by Mrs Joëlle Elvinger, Member of the Court of Auditors, in Luxembourg on 4 May 2022.

Annexes

Annex I – Tenure of agricultural land (EU-27, 2016)

Abbreviations

AFCOS: Anti-fraud coordination service

AMS: Area monitoring system

CAP: Common agricultural policy

COCOLAF: Advisory Committee for the Coordination of Fraud Prevention

COSO: Committee of Sponsoring Organizations of the Treadway Commission

DG AGRI: Directorate-General for Agriculture and Rural Development

EPPO: European Public Prosecutor’s Office

GSAA: Geospatial aid application

IACS: Integrated administration and control system

IMS: Irregularity management system

LPIS: Land parcel identification system

OLAF: European Anti-fraud Office

PIF: Protection of the European Union’s financial interests, from the French “protection des intérêts financiers”

SMEs: Small and medium-sized enterprises

VAT: Value added tax

Glossary

Action plan: A document establishing the steps to take to achieve a particular goal

Advisory Committee for the Coordination of Fraud Prevention: OLAF body that coordinates how the Commission and Member States combat fraud affecting the EU’s financial interests.

Anti-fraud coordination service: Body designated by each Member State to facilitate cooperation with OLAF.

Arachne database: Data mining tool developed by the Commission to support managing authorities and paying agencies in the management and control of the ESI funds.

Area monitoring system: Technology for the systematic observation, tracking and assessment of agricultural activities using satellite data.

Artificial intelligence: Using computers to simulate human intelligence through capabilities such as learning and problem-solving.

Beneficiary: A natural or legal person receiving a grant or loan from the EU budget.

Big data: The processing, collection, storage and analysis of large amounts of unstructured data, offering the potential to use the resulting information for new insights.

Certification body: For agricultural spending, a public or private entity designated by a Member State to certify annually the reliability of the accounts of paying agencies, the legality and regularity of the expenditure and the proper functioning of their internal control systems.

Checks by monitoring: A substitute for on-the-spot checks, involving the systematic observation, tracking and assessment of eligibility criteria and obligations using satellite data.

Corruption: Abuse of public, corporate or personal power for illicit gain.

Digitalisation: The shift towards incorporating and using digital technology and digitised information to make processes and tasks simpler, faster, more efficient and/or more economic.

Direct payments: Support payments, mostly area-related aid, made directly to farmers under the European Agricultural Guarantee Fund.

Error: The result of an incorrect calculation or an irregularity arising from non-compliance with legal and contractual requirements.

Established fraud: An irregularity that a court of law has ruled to be fraud.

Fraud: Intentional and unlawful use of deception to gain material advantage by depriving another party of property or money.

Geospatial aid application: An online tool for submitting area-based claims for agricultural aid.

Integrated administration and control system: EU mechanism used by Member States to manage and check payments made to farmers under the common agricultural policy.

Irregularity management system: Application that Member States use to report irregularities, including suspected fraud, to OLAF. 

Irregularity: An infringement of EU (or relevant national) rules or contractual obligations.

Land parcel identification system: A database of agricultural land in the Member States, used in the payment of direct aid under the common agricultural policy and in eligibility checks on farmers' claims.

Machine learning: The process in which an IT application uses artificial intelligence to improve its performance on a specific task.

Market measure: Public intervention in the agricultural markets to mitigate the effects of price drops and structural difficulties through sector-specific support (e.g. fruit and vegetables, wine, school milk).

Paying agency: Body accredited by a Member State to manage and control EU agricultural spending.

Red flag: Risk indicator that a transaction or other activity could be fraudulent.

Rural development programme: Set of national or regional multiannual objectives and actions, approved by the Commission, for the implementation of EU rural development policy.

Shared management: A method of spending the EU budget in which, in contrast to direct management, the Commission delegates to the Member State while retaining ultimate responsibility.

Small and medium-sized enterprises: A size definition applied to companies and other organisations, based on the number of staff employed and certain financial criteria. Small enterprises have fewer than 50 staff, and turnover or a balance sheet total not exceeding €10 million. Medium-sized enterprises employ fewer than 250 staff, and have turnover up to €50 million or a balance sheet total up to €43 million.

Statement of assurance: A statement published in the ECA's annual report, setting out its audit opinion on the reliability of the EU accounts and the regularity of the transactions which underlie them.

Suspected fraud: An irregularity that gives rise to administrative or judicial proceedings to establish whether it was fraudulent.

Replies of the Commission

https://www.eca.europa.eu/en/Pages/DocItem.aspx?did=61337

Timeline

https://www.eca.europa.eu/en/Pages/DocItem.aspx?did=61337

Audit team

The ECA’s special reports set out the results of its audits of EU policies and programmes, or of management-related topics from specific budgetary areas. The ECA selects and designs these audit tasks to be of maximum impact by considering the risks to performance or compliance, the level of income or spending involved, forthcoming developments and political and public interest.

This performance audit was carried out by Audit Chamber I Sustainable use of natural resources, headed by ECA Member Joëlle Elvinger. The audit was led by ECA Member João Figueiredo, then taken over by ECA Member Nikolaos Milionis, supported by Paula Betencourt and Kristian Sniter, Heads of Private Office and Matteo Tartaggia, Private Office Attaché; Richard Hardy, Principal Manager; Michela Lanzutti, Head of Task; Antonio Caruda Ruiz, Servane De Becdelievre, Jan Machán, Adrien Meric and Milan Šmíd, Auditors. Marika Meisenzahl provided graphical support.

 

From left to right: Antonio Caruda Ruiz, Michela Lanzutti, Matteo Tartaggia, Nikolaos Milionis, Kristian Sniter, Servane De Becdelievre, Jan Machán, Marika Meisenzahl.

Endnotes

¹ Article 3 of Directive (EU) 2017/1371 (‘PIF Directive’). The abbreviation PIF comes from the French ‘protection des intérêts financiers’ and refers to the protection of the EU’s financial interests.

² See e.g. the Association of Certified Fraud Examiners.

³ The PIF reports are available on the Commission’s website.

⁴ Belgium, Bulgaria, Czechia, Germany, Estonia, Greece, Spain, France, Croatia, Italy, Cyprus, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Austria, Portugal, Romania, Slovenia, Slovakia and Finland.

⁵ Article 7(3)(b) of Regulation (EU) No 1306/2013; Annex I of Commission Implementing Regulation (EU) No 908/2014.

⁶ Special reports 01/2019 ‘Fighting fraud in EU spending’, and 06/2019 ‘Tackling fraud in EU cohesion spending’.

⁷ 2019 annual report, paragraphs 6.34-6.41 and 6.44.

⁸ Commission’s 2020 PIF report, Annex 1, COM(2021)578.

⁹ Special reports 01/2019 ‘Fighting fraud in EU spending’, paragraphs 23-28 and 06/2019 ‘Tackling fraud in EU cohesion spending’, paragraphs 47-57.

¹⁰ DG AGRI Anti-fraud strategy, version 4.0 – Ares(2020)5099349, paragraph 7.3.1.

¹¹ 2019 annual report, paragraphs 6.35 and 6.36.

¹² 2016 annual report, Annex 7.3, example 1.

¹³ 2018 annual report, Box 7.3.

¹⁴ OLAF report 2018, p. 28.

¹⁵ 2017 annual report, Box 7.6.

¹⁶ Case reported in IMS database.

¹⁷ Transaction audited for the 2016 Statement of Assurance.

¹⁸ Articles 33(1) and 36(5) of Regulation (EU) No 1307/2013.

¹⁹ European Parliament (EP), ‘Extent of farmland grabbing in the EU’, p. 15; Transnational Institute (TNI)-European Coordination Via Campesina (ECVC), ‘Land concentration, land grabbing and people’s struggles in Europe’, p. 16.

²⁰ EP, ‘Addressing the human rights impacts of ‘land grabbing’’.

²¹ EP, ‘Extent of farmland grabbing in the EU’.

²² OLAF report 2017, pp. 20-21.

²³ OLAF press release No 03/2021.

²⁴ Judgment of 17 December 2020, Land Berlin, Case C-216/19, paragraph 34.

²⁵ Articles 58(2) and 59(1) of Regulation (EU) No 1306/2013.

²⁶ Judgment of 24 June 2010, Luigi Pontini and Others, Case C-375/08, paragraph 90.

²⁷ Article 15(2) of Delegated regulation (EU) No 639/2014; Case C-216/19, paragraph 42-43.

²⁸ Case C-216/19, paragraph 45.

²⁹ DG AGRI Note on the requirement of ‘eligible hectares at the farmer’s disposal’ (DS/CDP/2021/08).

³⁰ The OLAF report 2018, pp. 27-28.

³¹ COSO, Fraud risk management guide, p. ix.

³² DG AGRI Anti-fraud strategy, version 4.0 – Ares(2020)5099349, pp. 22-26.

³³ Annex I to Regulation (EU) No 907/2014.

³⁴ Answer given by the Commissioner to parliamentary question P-004224/2020 on 27 August 2020.

³⁵ Special report 04/2020, paragraphs 11-12 and 16-18.

³⁶ Special report 04/2020, paragraphs 17-18.

³⁷ Article 59(2) of Regulation (EU) 2021/2116.

³⁸ Statements on Regulation (EU) 2021/2116 (2021/C 488/02).

³⁹ Special report 04/2020, paragraph 82.

⁴⁰ DG AGRI annual activity report 2020, Annex 2, p. 25 (result indicator 3.5).

⁴¹ Paying agencies from Estonia, Croatia, Italy, Lithuania, Romania, Slovenia and Slovakia.

⁴² Paying agencies from Greece and Spain (3).

⁴³ Paying agencies from Belgium, Bulgaria, Spain (2), France, Italy, Luxembourg, Poland, Portugal and Sweden.

⁴⁴ The Largest 50 Beneficiaries in each EU Member State of CAP and Cohesion Funds.

⁴⁵ Requirements for a single database of beneficiaries.

⁴⁶ Commission’s 2020 PIF report, p. 44.

⁴⁷ Article 69(2) of Regulation (EU) 2021/1060.

⁴⁸ Article 59(4) of Regulation (EU) 2021/2116.

Contact

EUROPEAN COURT OF AUDITORS
12, rue Alcide De Gasperi
1615 Luxembourg
LUXEMBOURG

Tel. +352 4398-1
Enquiries: eca.europa.eu/en/Pages/ContactForm.aspx
Website: eca.europa.eu
Twitter: @EUAuditors

More information on the European Union is available on the internet (https://europa.eu).

Luxembourg: Publications Office of the European Union, 2022

PDF	ISBN 978-92-847-8010-5	ISSN 1977-5679	doi:10.2865/52418	QJ-AB-22-011-EN-N
HTML	ISBN 978-92-847-7989-5	ISSN 1977-5679	doi:10.2865/944695	QJ-AB-22-011-EN-Q

COPYRIGHT

© European Union, 2022

The reuse policy of the European Court of Auditors (ECA) is set out in ECA Decision No 6-2019 on the open data policy and the reuse of documents.

Unless otherwise indicated (e.g. in individual copyright notices), ECA content owned by the EU is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence. As a general rule, therefore, reuse is authorised provided appropriate credit is given and any changes are indicated. Those reusing ECA content must not distort the original meaning or message. The ECA shall not be liable for any consequences of reuse.

Additional permission must be obtained if specific content depicts identifiable private individuals, e.g. in pictures of ECA staff, or includes third-party works.

Where such permission is obtained, it shall cancel and replace the above-mentioned general permission and shall clearly state any restrictions on use.

To use or reproduce content that is not owned by the EU, it may be necessary to seek permission directly from the copyright holders.

Software or documents covered by industrial property rights, such as patents, trademarks, registered designs, logos and names, are excluded from the ECA’s reuse policy.

The European Union’s family of institutional websites, within the europa.eu domain, provides links to third-party sites. Since the ECA has no control over these, you are encouraged to review their privacy and copyright policies.

Use of the ECA logo

The ECA logo must not be used without the ECA’s prior consent.

GETTING IN TOUCH WITH THE EU

In person
All over the European Union there are hundreds of Europe Direct Information Centres. You can find the address of the centre nearest you at: https://europa.eu/european-union/contact_en

On the phone or by e-mail
Europe Direct is a service that your questions about the European Union. You can contact this service

• by freephone: 00 800 6 7 8 9 10 11 (certain operators may charge for these calls),
• at the following standard number: +32 22999696 or
• by electronic mail via: https://europa.eu/european-union/contact_en

FINDING INFORMATION ABOUT THE EU

Online
Information about the European Union in all the official languages of the EU is available on the Europa website at: https://europa.eu/european-union/index_en

EU Publications
You can download or order free and priced EU publications at: https://op.europa.eu/en/web/general-publications/publications. Multiple copies of free publications may be obtained by contacting Europe Direct or your local information centre (see https://europa.eu/european-union/contact_en)

EU law and related documents
For access to legal information from the EU, including all EU law since 1951 in all the official language versions, go to EUR-Lex at: https://eur-lex.europa.eu/homepage.html?locale=en

Open data from the EU
The EU Open Data Portal (https://data.europa.eu/en) provides access to datasets from the EU. Data can be downloaded and reused for free, both for commercial and non-commercial purposes.