This week's opinion: 12 November 2020

I am always interested in statistics and how they should, and should not, be interpreted. 

For example in the Individuals, Small Business and Agents Customer Survey 2019 recently published on HMRC’s website (tinyurl.com/yyj6j9fj), there is an astonishing figure on page 132: 6% of agents thought that tax evasion was acceptable in some or all circumstances – 1% did not know. Can that really be true? If there are agents out there who think that way, are they really going to say that to somebody researching on behalf of HMRC? 

Surely what is really happening is people are either deliberately trying to wind up the researchers, or else not really listening to the question and giving any answer just to get the survey over. I know from experience that I start with the intention of giving thoughtful and considered answers but, well before it has finished, I mentally switch off and am prone to say anything that comes into my head. 

The other figure is perhaps more significant. For the question ‘HMRC gets transactions right?’ 77% of small business agree whereas only 57% of agents agree. What does that mean? Does HMRC really do better with smaller businesses? It might, but is there another way of looking at it? 

Is the reality that there is no difference, but that agents are more likely than small businesses to know whether HMRC has got a transaction right? If so, that statistic is telling us more about agents’ knowledge of the tax system than it is about whether HMRC gets things right. That might be valuable in its own right but it is not necessarily what it might appear to do at first instance.