Does anyone understand how the retail sales measures manage to come up with completely different outcomes?

One of the analysts - I think it was the late lamented Richard Ratner, but I may be wrong - used to refer to life on what he called “Planet ONS”, a place where the retail landscape was completely different to here in the UK. I thought of that this morning when I saw the official sales figures for July, which claimed that the volume of sales through food stores fell in the month, while non-food sales were up, and those in “non-specialised stores”, which I can only assume are department stores, soared nearly 10%.

None of that sounds all that right to me, and the sense of the figures being a bit incongruous was only heightened by looking at the latest BRC figures, which claimed that over the three months from May to July, food sales grew at roughly double the rate of non-food. So who’s right? Any of them?

The problem with the ONS figures is that the methodology is so complex that it’s impossible for a mere mortal to get their head around, but sometimes with statistics the process can be so thorough you end up with a picture which is removed from reality. Indeed not that long ago it changed the methodology it used, in what looked like a tacit admission that the figures had previously been inaccurate.

A lot of the retailers I speak to are sceptical about whether any of the indices really give an accurate reflection of the market. Why does it matter? Well the ONS stats are used by policy makers as a barometer of retail spending. But if retailers aren’t confident that they’re accurate, what hope can the government or MPC have of making the right economic decisions based on them?