Bill Jamieson: Why our GDP numbers are the hottest show in town

Published on Thursday 28 October 2010 12:17

LET'S hear it for Mince! I refer to that longest-running St Andrew's House show playing to packed houses - and with different musical numbers every other month.

Don't let the ironically arcane subtitle put you off. Scottish Quarterly GDP statistics is a not-to-be-missed romp. Not even Creative Scotland or the National Theatre could come up with a production that so convincingly shocks and surprises at every turn.

They may keep you waiting five hours in the foyer to build up the anticipation. But there's not a dry eye in the house by the end. The big numbers never cease to surprise, from the opening floodlit chorus-line - "Flash Estimate" - to the big full-orchestra theme, "Quarterly GDP''. Then there's that long toe-tapping tail of post-interval knockouts: "Final Version" and "Statistical Revisions".

Clap? Your hands will be in bandages by the end.

You may remember last week I referred here to the new number introduced to the second-quarter show: Scotland's Knockout Construction Boom. This had the whole house clapping as the scantily clad showgirls came on stage with cards showing an unbelievable number for construction sector output in the March-June quarter: "Up 10.4 Per Cent!" It was this stunner that lifted Scotland's overall second-quarter GDP gain to 1.3 per cent.

Now, sitting in the front stalls with my bow tie and review jotter, I was finding it hard not to join in and clap. But then I spotted Michael Levack, chief executive of the Scottish Building Federation, just a few seats along. And I saw something odd. Michael wasn't clapping at all. In fact, he was bent double, his two hands busy on the floor trying to lift his jaw back into position. "Construction boom?" he kept gasping as I tried to steady him with a stiffener at the interval. "Where's the construction boom?"

The two most popular explanations were that the figures reflected a bounce from notably severe weather in the first quarter, then there may have been the local authority rush to get work started before the end-year budget shutters came down.

However, I had missed an extra paragraph in the four pages of small print warnings that come with the statistical release.The normal pro forma paragraphs were all there: "Scottish GVA estimates will generally be less reliable than the equivalent estimates for the UK..." and "Furthermore, the survey data for Scotland tend to be based on smaller numbers of units, making figures for Scotland more likely to be susceptible to statistical variance..."

But now there was an extra one: "There are substantial historical revisions in this publication," it read, "due to the methodological changes detailed in Note 5, most significantly due to benchmarking of GVA data to the 1998-2007 time series of Supply & Use tables in the Scottish Input-Output analysis". After this mind-numbing explanation there was reference to a separate methodology paper, which few by this time will have had the will to read.

We are indebted, therefore, to John McLaren of the Centre for Public Policy for Regions who did read on, and this is what he found. The data for construction has been "re-profiled", including downward revisions of between 11 and 13 per cent between 2003 and 2005. This left a profile for construction growth of 8 per cent between 1995 and 2005 (ie, less than 1 per cent a year) followed by a 17 per cent jump in the next two years to 2007.

By way of comparison, the previous figures showed growth of 22 per cent between 1995 and 2005, followed by 2 per cent growth over 2005-2007 - a markedly different picture to the one we had been previously given.

Figures for the UK as a whole showed construction rising by 27 per cent between 1995 and 2005 and by 3 per cent from 2005-2007 - again, a remarkably different picture to Scotland's new profile.

McLaren says: "I can make little sense of the new profile for Scotland." Nor, I suspect, can anyone.

Similar enormous changes have been made, he points out, in other areas. For example, the retail and wholesale sector has been revised down 14 percent in 1998; electricity, gas and water down 44 per cent in 2000; drink down 19 per cent in 2001; transport equipment up 33 cent in 1998; textiles down around 50 per cent over 1998 to 2000. This re-benchmarking suggests less a fine nuancing and fine-tuning of data than calling out utterly random new numbers from some demented tombola.

No explanation is proferred as to the "meaning" of the changes and the extent to which they related to economic conditions at any given time. So what are we left to deduce? A mystified McLaren is driven to two explanations. One is that "the old data was awful and told us nothing about the true path of the economy in many sectors". The other explanation is radically different: "The new data is awful and simply confuses us over what the true path of the economy has been in many sectors post-1995."

And all this matters, because it is the critical foundation and framework for economic policy decisions and resource allocation.If the data is unreliable, what chance then of the policy constructs that come to rest upon it?

Similar revisions - though not as dramatic - have affected the much better than expected "flash estimate" of UK third-quarter GDP released last week.

This showed that while the economy slowed from its 1.2 per cent growth pace in the second quarter, the initial estimate for the July-September period of 0.8 per cent was twice the consensus expectation.

As with the Scottish GDP numbers for the second quarter, the UK figure was flattered by a 4 per cent leap in construction output compared with the immediate previous three months. This is likely to relate to public sector work. But might the figures be underestimating inflation and thus overestimating output?

Is the UK really enjoying a better-than-expected recovery - or were last week's GDP numbers a "one-off"? For those who have argued that there is a huge "output gap" in the economy, does this now mean it has narrowed? How is the Bank of England's Monetary Policy Committee likely to react at its meeting this week? A resort to QE2 had come to look increasingly likely.

Now the committee's task looks altogether more difficult. If it does nothing, it stands to be accused of complacent dither as a growth recession takes hold. If it opts to embark on another bout of quantitative easing it may have lit the fuse for a recovery-destroying inflation surge 18 months to two years ahead from here - earlier if that output gap has indeed narrowed.

The worry is that policy is being steered by figures that reflect significant statistical revision rather than a genuine recovery story, with disastrous implications for policy. Caution would seem to be called for now.

As for Scotland's GDP, I have complained about its timeliness and oddity so often that I am now tempted to throw in the towel, summon a doctor for Mr Levack and return to the front stalls for the rest of Mince! All we need now is Lady Gaga as lead vocalist.

In my youth, when it was all Bill Haley, Fats Domino and Jerry Lee Lewis, we used to jive down the aisles at this sort of thing. Mince! may be a magnificent retro throwback.