Mr. Kevin Cardiff:

It is a pleasure to be in Dublin to present the annual report of the European Court of Auditors. We sent the annual report itself to the committee when it was published in November, and we have also sent to the secretariat our information note consisting of a summary of the key points, which members might get a chance to browse through at some stage. We have a slide presentation today, which has been circulated, and I propose to give my opening remarks with that presentation. Do not fret, Chairman; the second set of slides are only for supplementary information, so I will not be trying to get through them all.

The European Court of Auditors is an organisation based in Luxembourg with around 900 people, including auditors, translators, administrators and support staff, carrying out audits and producing audit reports and opinions. We produce around 70 reports per year. With some limitations, we are the independent external auditor for the EU and its institutions and agencies, so many of those 70 reports are short reports on individual EU agencies. Our annual report is the key report each year on the legality and regularity of EU spending, and producing that report takes up about half of our total resources. I am here today principally to deal with that. However, we also produce many special reports, dealing with economy, efficiency and effectiveness. These are the European equivalent of the VFM audits from the Comptroller and Auditor General with which committee members here are familiar. These relate to selected areas of EU spending or operations.

At the moment, EU expenditure amounts in cash terms to about €136 billion per annum. In other words, it is about the same budget as a medium-sized European member state. Agriculture and rural development, environment, fisheries and health - generally known as the natural resources area - taken together amount to more than 40% of the budget. The Common Agricultural Policy is probably the single biggest policy area in terms of spending. Regional policy and so on amounts to about 30%, so these are the two big areas. The slide in front of us shows the breakdown of that spending.

Ireland makes an annual contribution to the spending in the same way as every other member state, based on the same calculation methods. For Ireland, that is about €1.2 billion for 2012. The next slide shows that although we have paid around €1.2 billion, Ireland received around €2 billion one way or another, so that is a positive difference of around €600 million. The slide also shows the areas in which Ireland received funds from the EU budget. The bulk of that funding is under the natural resources heading; in other words, principally for agriculture. As far as the European Court of Auditors is concerned, this is not money given to Ireland, but money used in Ireland for EU policy purposes. EU money is spent in a variety of ways and there are particular mechanisms used for getting the money out the door. The next slide shows that the bulk of funds are disbursed under the overall heading of "shared management". In other words, the moneys are handed by the EU to the member states and their regional authorities - in some countries there are many regional authorities - so the management of the funds is shared between the European Commission and the member states. In truth, it is mostly in the hands of the member states, and therefore it is for the member states themselves to ensure that the funds go to the right recipients for the right purposes at the right time. Our audits look at both systems and transactions at the Commission level but also at member state level and even regional level.

The Commission itself runs some programmes directly. These are known as direct management programmes and about 20% of the spending goes through them. In those cases, our audits look at the Commission's systems. I will skip the slide dealing with indirect management, which is about 5% of spending. The next slide shows where the management controls lie. In direct management areas, management is with the Commission, which initiates the transactions, verifies them and pays them, whereas in the case of shared management the system is quite a bit more complex. The Commission still has top-level responsibility, but typically there will be several layers of management at the member state level. For example, in some member states which are both heavily decentralised and also federal in nature, there could be quite a few different paying agencies at the regional level, each reporting to the state governments, which report in turn to the national-level governments. So while there are 28 member states, there are more than 80 paying agencies responsible for disbursing EU funds, and each agency reports to national level bodies and is overseen by external certifying bodies. Slide 26 in the supplementary material describes those control systems in more detail, and committee members can look at that slide at a later stage if they wish, but we can see that the system is complex.

The next slide shows what we call the statement of assurance exercise, presented in our annual report. We are interested in two key things. First, are the financial statements fair and accurate statements of the financial position? In other words, are the accounts more or less accurate? The answer to that is more or less always "Yes". Second, are the EU revenue and expenditure transactions carried out in a legal and regular way? In other words, do they all comply with the rules?

The answer is much more qualified - it is certainly not "Yes". Members will note from the picture shown on this slide that the ECA audit is literally a boots-on-the-ground affair. There is a huge amount of documentary work but we send auditors literally into the field - into the factories - in many places around the world because EU money is disbursed to many places around the world.

In the case of this year's annual report, which the members are kind enough to let me present, while they can take it that, as far as we are concerned, the EU accounts are within very tight ranges more or less accurate, the payments contain many errors, well in excess of what we regard as a material level of error - in other words, a statistically noticeable level of error. This is shown in slide 9. This year we calculate an error rate, on a very arcane and complex basis, of around 5%, which, as the Chairman said, represented an increase compared to the 2011 level. There are some methodological changes, but the impact of those is small and the error rate seems to be on a consistently, if not a dramatic, upward trend. Error rates had fallen in the period up to 2009 but have been rising relatively gently since. Members can note that from the graphic on page 10. There had been considerable improvements between the 2006 and 2009 annual reports but there has been some disimprovement since. There are probably some cyclical effects at work here, which we can discuss later, but the disimprovement seems real. We think it is not just a statistical artifact. To be clear on this, the error rate is not a calculation of how often there is an error, because there are errors in some areas in more than half of the cases. The rate would therefore be much higher if it was simply the occurrence of errors. This is the impact of errors on the budget. In other words, if, for example, there was a 2% overpayment, that error would be counted as having a 2% impact. It is also not the same as saying that money was wasted. If, for example, a beneficiary makes an error in a procurement procedure or if money is disbursed to a school in the wrong region, one would say that was in breach of the rules - namely, a payment that might not have been due to be made - yet one would not say necessarily that the money was wasted. It might still be used for a valid and positive public interest purpose. In other cases - thankfully, we have come across relatively few of these - there would be fraud or an over-claim, and in those cases the money is definitely wasted. Our error rate does not reflect fraud. While we come across some fraud cases in our work, those are relatively few. In practice, in so far as our auditors are concerned, most of the errors they find are actually that - errors - rather than fraud. That means that some of what is written in the press about the European Court of Auditors saying that €5 billion was wasted last year is not accurate. When we do our performance audits, which is where we examine the effectiveness and value-for-money considerations, we might find that even more money has been wasted than the 5% error rate would suggest, but that is a different matter. This is just about error.

Moving on to slide 11, members will note the different types of error that arise. A large proportion of them are serious errors in public procurement procedures. When EU money is disbursed it is supposed to done in a way that is fair and leaves the market open and so forth. There are quite of lot of errors, but the rules are quite complex in that area. Ineligible projects are another large problem. Another is ineligible costs; a project is eligible but, to give an example, staff costs might have been claimed where they should not have been claimed. These things arise relatively frequently also. There is also a large body of errors which are of very little interest. They arise often - minor clerical errors and the like - but they do not have an impact on the budget. It is important to note that when we say there is a material level of error in the transactions we are not counting every one of those simple errors as if the whole transaction was invalid. What we are saying is that impact of the error on the budget is what is reflected in our error rate. We usually do not put down the whole transaction as being flawed if there is a small mistake in it.

The most error-prone spending areas - it has been this way for quite a while, as the members will note on page 12 - are in natural resources spending areas such as rural development, excluding direct support to agriculture, and the regional policy areas. That is not out of line with previous years. In some ways it is not out of line with what one might expect intuitively. If a farmer is receiving his single payment in a year, that is a much simpler transaction than if he is receiving support under an environmental heading and he has to prove various environmental conditions have been met and so forth. In some ways, the simpler the scheme, the lower the level of error. That is intuitively correct. What is slightly worrying from our point of view is that in the shared management areas, most of the errors that our auditors find could reasonably have been spotted and corrected by the member state itself. It is an important point for us because it means that to the extent that controls operate, they need to operate close to the payment rather than at the central level.

Since it is biggest area of spending for Ireland, I will deal with agriculture as a key example on slide 14. It shows some of the results in the market and direct support areas - in other words, not the regional development area. We identified that there was a material level of error in the EU expenditure in this area of around 4%. This is considerably less than the error rate for rural development, but the schemes are simpler and, by definition, more direct. The most frequent error is over-declaration of land areas. Most of the payments are tied to land area. Those over-declarations can be a simple error, they can be just misunderstandings or, occasionally, they can be deliberate.