Skip to main content
European Commission logo print header

Statistical Methodologies for Auditing CAP Expenditure and Control of Fraud against the Budget of the EU

Cel

Specific Objectives:
1. (WP1_T1). To define audit sources and audit information to be stored per source using the following three categories of sources of audit information: reports of certifying bodies, enquires of the Audits of Agricultural Expenditure Units of the Commission as summarized in ad- hoc clearance decisions and Court of Auditors annual MUS exercises. Working in collaboration with Audit of Agricultural Expenditure Directorate, the action team will formulate available audit sources and, for each source, the audit information to be stored in a database that interested and acceptable parties could use for audit findings relative to CAP. Duration: 6 months;
2. (WP1_T2) To carry out data retrieval from selected sources and storage into the database. In year 2003, one source from each of the three source types mentioned in WP1_T1 will be retrieved and stored into the database. Duration: 12 months.
3. (WP2_T1). To monitor measures with a view to overpayment through calculation and tabulation of summary and descriptive statistics. These estimates will be calculated on the data stored for one source for each category of sources mentioned. This task will not be concluded in 2003. Duration: 3 months.
4. (WP2_T2). To extend audit planning based on risk analysis considerations on the basis of estimates budget fractions reportedly paid irregularly for small samples. This task will not be concluded in 2003. Duration: 3 months.
5.(WP4). To further develop ARIADNE application to include the user-friendly interfaces for the procedures to data-mine large two- way tables of counts and tables of totals for continuous variables, descriptive summary statistics for categorical and numerical variables, data dictionaries for data bases with various filtering options. Duration: 12 months Anticipated milestones and schedule Deliverable for WP1_T1 - end June 2003 Application ARIADNE, data mining tools for large 2-way tables of counts & totals of continuous variables - end June 2003 Application ARIADNE, compilation of summary and descriptive statistics for numerical and categorical variables - end September 2003 Deliverable for WP1_T2 - December 2003.
Planned Deliverables - 1st Deliverable (WP1_T1). Report on audit information to be stored per each audit source. - 2nd Deliverable (WP1_T2). Data retrieval and storage of one source per each source category. Storage of audit data for one source from each source category. - 3rd Deliverable (WP4).Application ARIADNE, year 2003.
Summary of the Action:
The aim of this action is to develop and demonstrate an integrated methodology for auditing expenditure for the Common Agricultural Policy (CAP) in a cooperative mode between European institutions and the Member States. Research will be conducted into more economical ways for statistical auditing to estimate overpayment in the CAP, by using previously available audit results to achieve higher precision for the same sample sizes or smaller sample sizes for the same precision. In close cooperation with Audit of Agricultural Expenditure Units, the action team will store results of audits of agricultural expenditure in a database of audit results (WP1). The exploitation of database of audits of agricultural expenditure (WP2) includes monitoring measures with a view to overpayment; audit planning based on risk analysis considerations; and use of the database of audit results to enhance of statistical exploitation for each of the sources of the database. Sources for strategic intelligence in antifraud applications with regard to the protection of the budget of the European Union are payments to beneficiaries, data on trade, audit findings of control or certifying bodies when appropriately documented, and, most importantly, databases of cases of irregularity or fraud detected and reported by Member State authorities. The proposed action gives particular emphasis to the latter databases but also pertains to all data sources mentioned above, particularly when used jointly. The proposed action comprises three additional work packages. Applied statistical research work in mining databases for antifraud applications (WP3) will address the problems of enhancing quality of stored data.

Results obtained, tested and pertinent in mining databases for antifraud applications will be accessible for relatively easy use from an application of user-friendly interfaces (WP4). Results interfaced will be evaluated by independent statistical experts of an academic profile and confirmed to be domain-relevant by subject matter experts. Finally, (WP5), evaluation and dissemination of results of the four previous work packages will be made by presentations to academic audiences, user DG and other similar audiences, training of staff of user DGs and their partner organizations. In 2003, tasks of WPs 1 2 and 4 will be completed as follows: WP1_T1. Definition of audit sources and audit information to be stored per source. WP1_T2. Data retrieval from selected sources and storage into the database will be completed. The following tasks of WPs 2 and 4 will also be addressed. WP2_T1. Monitoring measures with a view to overpayment. WP2_T2. Audit planning based on risk analysis considerations. WP4. Building user-friendly interfaces for tested, debugged and accepted programs for extracting knowledge from databases, (application ARIADNE). Rationale Currently, payments to beneficiaries of the Common Agricultural Policy (CAP) are made via paying agencies, organizations that are submitted to annual and ad- hoc internal and external audits and controls. The proposed action aims to develop an integrated methodology for co-operative auditing so as to make estimates of overpayment more efficient and corrections for overpayment more acceptable, for the benefit of European Institutions, Member States and European citizens.

The action also aims to contribute to fraud prevention by continuing the development of statistical methods for the extraction of knowledge from databases of committed, detected and communicated cases of fraud and irregularities. The JRC is independent from commercial or national interests and therefore its conclusions and findings are beyond any doubt, even when the questions addressed as to how to audit, how to estimate overpayment and therefore how to correct each Member State for overpayment, involve directly national interests or commercial interests, when the Certifying Bodies are a part of the private sector.

The JRC guarantees the protection of the confidentiality of the data entrusted to it by other services of the Commission both in the storage of raw data and in appropriate masking in the presentation of results for evaluation by third parties. Being a service of the Commission, the JRC has support, documentation, knowledge, and facilitated access to a host of Commission databases. The special position of the JRC within the Commission has enabled it in the past to confederate different databases and produced results of particular interest. Last but not least, the JRC has demonstrated considerable and convincing technical and topical competence in statistical estimation problems arising in risk analysis, in sampling and estimation for auditing, and in statistical data mining work quoted in a Commission communication. Work of the proposing team has also been presented to academic audiences or antifraud organizations, such as the EPFL statistics seminar in December 1998, and presentations in a University of Bologna workshop in October 2000, the Guardia di Finanza general staff in July 2001, etc.

Zaproszenie do składania wniosków

Data not available

Koordynator

Institute for the Protection and Security of the Citizen
Wkład UE
Brak danych
Adres
Via E. Fermi 1
I-21020 Ispra
Włochy

Zobacz na mapie

Koszt całkowity
Brak danych