On that theme, have a look at figures 26 and figures 26a of his report. It’s one of the first places I have looked at critically because that is where you can see the basis of the calculation of the £40bn “tax lost” number in one place:
Now, the EU forecast in figure 26 still includes bad debt and criminal attack.
But, it appears that the UK budget figure doesn’t. Here’s the preceding paragraph explaining why I think it doesn’t:
If he is using the Treasury figure as a proxy for the VAT gap less bad debt, then he’s already taken off bad debt in reaching that 9.5% in 2011/12. Or, at the very least, it is almost a figure that is the VAT gap less bad debt.
Now look at the fifth line of figure 26, “ad debt”. That 1.5% relates to the average bad debt from HMRC’s data for the previous five years.
The 9.5% already excludes bad debt. Richard is double counting bad debt against the UK budget figure when he removes it from the average..
So the numbers that he pumps into the third line of the next table, figure 26a are wrong.
Effectively, the VAT gap used in 26a is an average of HMT’s VAT gap including criminal attack (criminal attack won’t correlate to unrecorded sales) but excluding HMRC’s bad debt and the EU VAT gap including both criminal attack and HMRC’s bad debt, with HMRC’s bad debt taken off the average.
To be fair, I suppose it’s not double counting. It’s more like one-and-a-half counting…
Now, going back to the alternative, if Richard is not using HMT’s figure as a proxy for the VAT gap less bad debt, then he is using a dataset that he has no idea what it actually relates to.
I can’t decide which is less reassuring…
Either way, those calculations are a complete waste of space.