Sunday, 29 April 2012

How wrong does the ONS get its initial GDP figures?

After the Times' criticism of the Office for National Statistics for releasing an early estimate of the GDP data, that the newspaper didn't like, it is worth looking at how off these first releases tend to be.

Check out this graph

For the most part, the orange line is lower than the blue line, meaning there probably is some scope for the the figures to be revised up by the 0.2 percentage points. This would then mean that there would not have been two quarters of 'negative growth' or a recession to you and me.

Although somewhat amusingly, on of the few points where the initial estimate was higher was the one which prompted that gushing Times leader I flagged up in the last post, about how we had avoided double dip recession.

Saturday, 28 April 2012

Leveson gets political (and unattributed)

The Leveson Inquiry will apparently not expand its remit to consider Jeremy Hunt's handling of the NewsCorp/BSkyB affair in relation to the ministerial code.

A sensible move, given the inquiry was not set up with this in mind. After all it is an inquiry into the Culture Practices and Ethics of the Press, not of the cabinet. No doubt such an inquiry would prove just as intriguing, but for better or worse we don't have one of those.

So instead we are left wondering how the Government will respond, since Strategy A has apparently failed.  But what to make of Leveson's move to stop the Hunt issue going into the long grass?

Yes he has done what seems perfectly sensible, namely stopping his Inquiry being used to save a ministerial career.

But the way it has come out seems a little odd. The Daily Mirror for example are simply quoting an Inquiry Source rather than Lord Justice Leveson.

Could be that this is simply a way of making the official inquiry spokesman sound a bit more like a juicy confidential source. But it if this really is the view of the Leveson Inquiry why can't whoever it is just go on the record with their name, or better still have it come straight from Leveson himself?

Hopefully there isn't anything in it, but why give unattributed quotes to the Daily Mirror, the most anti-Coalition paper, and risk looking like the Inquiry is on their side?

The Hunt affair has risked turning the inquiry into something of a political play thing, with opposition MPs pressing the Inquiry not to become a get out clause for JH, and the Government pushing the other way. Whatever the Inquiry does it will unavoidably have party political implications.

Given the important and wide ranging issues the Inquiry is dealing with, Leveson needs to make sure his eventual report gets a fair hearing from both press and Government when ti finally reports. 

There is a delicate balancing act to be done between corners of the press which view the inquiry as a wheeze dreamt up by politicians to stop them scrutinising their behaviour too much, and embarrassing the Government to the point of antipathy to Leveson.

Perhaps you can't do one without the other, but it has to be hoped that Leveson maintains the Independence he has scrupulously shown so far, and doesn't leave his Inquiry open to any charges of having an agenda.

Other than better standards for the press, of course.

Update: The mystery continues over where that quote came from. Try Googling it. All I could find was PoliticsHome's reuse of the above tweet. It doesn't appear to have made it into the Mirror's coverage, nor any other papers' for that matter.

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.

Thursday, 19 April 2012

Does the media have a drink problem?

What is it with the media and binge drinking? No, not the stereotype of the ever thirsty roving reporter, but the coverage of drinking culture in England.

One need only do a quick Google trawl of newspapers such as The Sun or the Daily Mail, to see how much they love to regale us with reports of 'Binge Britain'.

But digging behind the source for today's Daily Mail report on the amount of lives that could be saved through a minimum price on alcohol, shows a glass half full, rather than completely empty.

What doesn't get as much coverage is that binge drinking in Britain has been on the decline. As shown in the graph, cribbed from a fantastic briefing on alcohol statistics from the House of Commons. (H/T Straight Statistics).

For men aged 16-24 it has been declining since 1998, a pretty long-term trend rather than one being related solely to penny pinching where pints are concerned in these tricky economic times.

Indeed, as pointed out by the King's Fund's John Appleby in a piece for the British Medical Journal, real terms spending has also been falling, save for a small rise in the last year.

So we are apparently binging less and spending less, so what's all the fuss about?

Well, hospital admissions related to alcohol have been on the rise. But rather than suggesting that the reduced amount of booze we are getting down us is actually worse for us, the time lag involved with some of the conditions (alcoholic liver disease, or mental problems associated with alcohol) suggest the admissions shouldn't necessarily mirror each binge drinking numbers.

But what about the louts and loutettes who, so the reporting goes will, be leaving a trail of destruction from town centres to A&E departments across the land this weekend?

Sadly the House of Commons paper only mentions that booze is a factor in 70 per cent of emergency attendances between midnight and 5am at weekend.

Sure enough this will be a drain on NHS resources, but what it doesn't tell us is the problem getting better or worse? More digging needs to be done, but will update with anything I find.