UK’s ONS gets retail figures wrong

OK, hands up. I have criticised the reliability of Chinese data regularly on this blog to the extent that I do not believe a lot of the published Chinese macro data.

Interestingly, I did wonder about the robust UK retail numbers that were coming out of the ONS in the UK. Surely UK statistics can be trusted. Surely.

Only now do we get the truth. How hard can it be – honestly. OVERSTATED by 56% – that is simply dreadful.

It is becoming clear that this financial crisis is great for getting the world back on a more realistic track – the party is over – we now need to clear up the mess left behind.

Heads should roll at the ONS. China’s statistical office can still learn a lot from this sort of misguided confidence that the head of the ONS showed in their own ability to collect a simple series of data.

There is political capital from posting wrong numbers whether it is to convince a domestic or overseas audience.

ONS gets sums wrong on retail sales [FT]

One of Britain’s most closely watched economic indicators has heavily overstated the quantity of high street sales over the past two years, the Office for National Statistics admitted on Friday.

Britain’s supplier of official statistics conceded that since the financial crisis began in August 2007, it has overstated the volume of retail sales growth by 56 per cent.

Many economists have been worried for some time that the published retail sales figures were too strong and have always received a furious response from the ONS.

Karen Dunnell, the national statistician, wrote to newspapers last October, insisting that “ONS retail statistics are the best available and are not inaccurate”.

She stuck to the same theme in another article, saying economists who had expressed surprise at the strength of ONS retail figures were upset because “City analysts also have a vested interest in not being proved wrong”.

Yet while the ONS head was defending the accuracy of the retail figures, industry experts knew the outdated nature of price measurement in the retail sales index was much more than a triviality.

The ONS previously said that between August 2007 and March 2009 retail sales volume grew 3.6 per cent. The changes announced on Friday mean it will now say the real rise in sales volumes was only 2.3 per cent.

Such a large difference in the one indicator that has persistently given a more positive account of Britain’s economy will cause red faces at the ONS, especially as it had insisted on the superiority of its retail data to unofficial estimates.

So confident has the ONS been that it warns users of the CBI or the British Retail Consortium sales data every month that these figures might not be “fit for purpose”.

The BRC on Friday welcomed the changes, saying they meant the official data would now be more in line with its figures.

The ONS’s old methodology failed to take sufficient account of goods that had risen strongly in price and so under-estimated the true rate of inflation in calculating the headline retail sales volume figures.

The ONS said on Friday it was changing the way it compiled its retail sales data to make sure it “more accurately captures recent trends in retail sales, including where consumers switch purchases to goods that have fallen in price’’.


Originally published at China Economics Blog and reproduced here with the author’s permission.