Mr. Derek Moran:

It is product by product rather than county by county. The brunt of it is taken out of bond. It gives us that level of insight.

VAT over the quarter is strong. Profiling is imperfect. We had the example of the pre-purchase of tobacco last year in advance of the introduction of plain packaging and we get a distortion and try to work our way around it. There is always the risk of profiling being wrong. Generally speaking, tax forecasting falls within a reasonable bound of statistical error. It is very easy to talk about the abstract. If it is on the negative, there are fiscal implications but it is a forecasting exercise. We are taking information now and predicting the future compared to the trends in the past. As Mr. Watt mentioned earlier, even GDP data is revised for five, six and seven years after the initial Estimates. We have to continually adjust for that. The point is to keep it continually under scrutiny and to review it in a root and branch way every few years. In terms of the elasticities, so to speak, of the ESRI, part of the joint research programme is to examine those very issues, give time to the analytical piece and look at the reason the income tax base is behaving slightly different across its components. Does Mr. McCarthy want to say anything on the forecasting side?