Mr. Kevin Cardiff:

This is the third meeting I have had with the committee, but my predecessors have also had meetings. We have the great pleasure as members of the court of trying to engage with every national parliament. Some parliaments do the engagement more or less like this, as a meeting with the EU committee of the relevant parliament, while others do not do it all. They do not think this is important for them.

We call ourselves the guardians of the EU finances, but of course we are just one part of an elaborate structure both for making payments from the EU budget and for protecting the budget from inappropriate use. The purpose of this session traditionally is to talk mostly about our annual report. The report is the court's assessment of the legality and regularity of payments during the year, but we also do a set of other reports. We do a specific annual report on each of the 40 or so EU agencies. We do special reports similar to the Comptroller and Auditor General's value for money, VFM, reports looking a little more at the effectiveness of the spend and not just the legality of it, and we give opinions on forthcoming EU legislation, usually when we are asked. This means in a typical year we might produce between 70 and 80 separate products, but of all of them, the annual report is the most significant and it takes approximately half our resources to produce.

Most EU spending in Ireland comes under the natural resources headings such as agriculture. The competitiveness in growth and jobs line in the budget also provides funding to Ireland. That includes a great deal of research spending. On the revenue side, Ireland's contribution is related to the size of the economy and VAT receipts. In 2014, for the first time, it became a net payer. There are different ways of calculating this. On some Commission calculations, that might be a year or two away yet, but basically we are a net payer country now, and from what one hears, that will begin to effect the policy mix in Ireland over the next few years. Traditionally, net payer countries are more demanding in terms of control structures, value for money and control limitation of the EU budget whereas, naturally, the recipient countries have an incentive to prefer an expansion of the budget.

The peak year for net receipts for Ireland from the budget was 1997, and it tailed off since as economic growth improved relative to other European countries, while 1991 was the peak in the receipts in percentage of GNP terms at 6%. In net terms this year, the EU spend in Ireland is more or less neutral but it was at one point 6% of GNP. When we needed it back then, we got quite a lot. Ireland's share of contributions to the EU budget is approximately 1% of the total. The six largest contributors account for three quarters of the revenue. Germany, France, Italy and the UK contribute 60% of the budget, and therefore they are the big payers. When we do our annual report, we take from this huge spend of €145 billion approximately 1,200 transactions. They are sampled on a representative basis more or less randomly within particular constraints. This means that because we sample by the euro, if a transaction is ten times as big as the next one, it has ten times as much chance of getting picked up in our sample. We try to make it representative. This means, though, that for small individual member states, the number of transactions selected would be relatively small.

That means we cannot make big, sweeping conclusions about the legality and regularity of transactions in Ireland because in the course of a year we might only sample six or eight, which is the appropriate proportion relative to the overall budget for Ireland. We give some data but it is not to be used to draw large conclusions. While we make great effort to ensure our sample of transactions is genuinely representative, for the smaller companies especially, the number of sample transactions is too small to make a country-by-country conclusion.

This year, as in all previous years in recent decades, we find the accounts of the European Union are true and fair and a proper representation of the Union's balance sheet, income and expenditure over the year. People say we do not sign off on the accounts but in fact, in private sector accounting, once the auditor has signed off on the "true and fair" criterion, the job is pretty much done. On that basis, the Union is getting a positive outcome from our audit. The treaty requires us to go much further than that and also to consider the legality and regularity of individual transactions, which is why we go into so much detail in the sample of individual transactions across the Union. We audit both the revenue and the expenditure of the EU but the revenue side is relatively simple - it is all between governments so it tends not to have big errors. This year again we found that the revenue side was free of material error. However, we found an error rate of 4.4% for expenditure, which is almost exactly the same as last year. Last year it was 4.5%, so allowing for sampling and so forth, that is pretty much exactly the same.

The EU has various own resources, that is, streams of income, but the total per member state is determined by reference to each member state's national income, GNI. This is by far the most important element of all the revenue calculations. While in the transactions themselves, revenue is free of material error, we draw attention in this annual report and previous special reports to the fact that there can be big changes in the GNI data. The transaction is based on the GNI figures, which are subject to revision. Those revisions can be quite large and they rely a lot on the quality of the product of the central statistics office of each relevant member state. At the end of 2014, there was a row about a sudden, supposedly unexpected demand for additional funds from the UK. What actually happened was that a known statistical problem which had been festering for some years was suddenly solved and the solution meant the UK had a bigger national income than it had previously reported which meant that the amount it had to pay was greater. There was a big rush to come up with a compromise solution that spread the payments. The biggest risk on the revenue side is not that the actual transactions will be wrong but rather that the underlying data will shift and change. If one looks at the Irish CSO data, the first two or three rounds of GDP, GNP or GNI data are always marked as provisional and often change as new information becomes available. The EU figures take account of the changes for previous years in this year's payment but obviously if there is a problem that has festered for many years, the amount can build up. We have looked at this in a special report and made some recommendations but the most important thing is that these reservations or festering problems are dealt with as early as possible.

The error rate on the spending figure of approximately 4.5% varies by type of spending. It is important to note the error rate is not an estimation of fraud or waste. I say it every year.