Mr. Niall Cody:

We have started the process. There was an initial discussion at board level on the review of the random audit programme. The random audit programme has been in place for a long number of years. It has run at approximately 400 cases per year, with cases selected randomly in the same way. This allows us to identify a trend over time. As a result of the changes we are making to our overall risk-based approach to interventions, now is an opportune time to review how we do it. A paper has been submitted to the board and we have asked for further work to be done. Obviously, before we sign off on any changes, we will discuss them with the Office of the Comptroller and Auditor General to ensure that we do not have a row in three years time when taxpayer compliance is reviewed and it says it would never have agreed to the changes.

Some of the work we are doing in the office concerns segmentation of our case base. The random audit programme is taken out of the 850,000 chargeable persons and companies that are self-assessment cases. However, there is a huge variety of self-assessment cases and the bottom end of the self-assessment register are those who are more akin to PAYE workers. For instance, they might have a foreign pension. They would be PAYE taxpayers if we had anyone to collect it from. We need to segment it.

I have examined some of the approaches. For example, HMRC in the neighbouring jurisdiction carries out 4,000 audits for its random audit programme but it takes three different sectors - sole traders, companies and SMEs - and leaves all the large cases out of the programme. We are interested in examining the random audit programme.

A couple of matters arise when this review takes place every few years. We always counsel against relying on the average yield. The average yield was hugely distorted by one case. A total of 79.9% of the cases have a yield of less than €2,000 and 1% of the cases had a yield in excess of €50,000. One can get a very false-----