The Public Accounts Committee has published a new report on HM Revenue and Custom's tax gap estimates this morning. As I will note  in a separate post, they are deeply critical, but not wholly for the same reasons that I am.

One issue that the PAC did not pick up on was that HMRC has admitted that their whole model for estimating the tax gap is flawed. I call that a big deal. I address the issue in this video, made before I saw the PAC report:

In summary, the case is a simple one, and one I have long made. HMRC has used a US model for estimating the tax gap. So, they sample some tax returns and find errors and then extrapolate that error rate across all the tax returns that they get.

The model works in the USA because everyone has to submit a tax return.

Only about a third of tax payers in the U.K. have to submit tax returns. And HMRC make the absurd assumption that those who do not submit tax returns - whether they are individuals, or the very large number (almost the majority) of companies that do not do so - make no mistakes in their returns as the have no income to declare.

That assumption makes no sense at all. In fact, not requiring returns is    a mechanism perfectly designed to help the evader and is bound to increase the tax gap.

The assumption HMRC has used by adopting a US model does not work then.

And that is why I have always suggested that there is very strong evidence that HMRC has substantially under-estimated the tax gap. I think it may be three times their own estimate, which is implausibly small and so persistently stated that its credibility is very low.