Friday 27 April 2012

The Times shoots the statistical messenger



Update:  The revised figures are in. As you have probably read by now, the ONS has downgraded its estimate down from -0.2 per cent growth to -0.3 per cent.

So rather than the preliminary estimate which the Times was so irked about being overly pessimistic, it the early figures painted a slightly too rosy view of where the economy is.

Perhaps The Times should take follow its own advice and wait until it has more information before passing judgement.

----------

The Times is thoroughly displeased with the Office for National Statistics.

What have our official number crunchers done to provoke the paper's ire? Well they made news of a double dip recession semi-official yesterday, and in a quintessential case of messenger shooting, an editorial today blames the poor old ONS for putting us in recession.

The Times asks "whether it is worth the Office for National Statistics rushing out estimates that are so often subject to big revisions. Few other experts think that the 0.2 per cent contraction is an accurate picture of the economy."

The paper is cheesed off that the latest figures show a 0.2% contraction in the last quarter, aka two quarters of negative growth, aka double dip recession, aka not a recovery.

The problem lies in the preliminary estimate which the ONS produces, the like of which was released yesterday. Rather than waiting for all the data an initial figure goes out which is revised a few weeks later. At such fine margins a few decimal points can change the whole narrative on the economy,

Should that revision send us from -0.2% to 0.0% change in GDP a double dip will have been avoided, and everything is great.

Couldn't the stats bods just have waited until they were absolutely really, really sure we were in recession? After all there are plenty of other economic groups who estimate we didn't go into recession. Why do we have to bother with this 'official' estimate when there are ones out there which fit our argument a bit better.

While there is a point here about how much use this preliminary estimates are if they keep on getting revised, getting annoyed at the statisticians when their figures don't say what you want is pretty cheap for  a subscriber newspaper.

Why is it bad for the ONS to 'rush out' figures but ok for other experts to pronounce on the health or otherwise of the economy in a timely fashion.

Especially when last year the paper was quick to jump on an ONS preliminary estimate as very good news, claiming the economy while not fully recovered had "started on that trajectory".

"A few weeks ago, commentators were rubbishing his strategy on the ground that it risked a double-dip recession. Preliminary figures released yesterday indicate that this at least was avoided"

Now these same preliminary figures show we are in a, revision pending, double-dip recession.

So early estimates are fine, but good news only please.

No comments: