Dodgy data

At the world's least ecologically sound event -- the Rio +20 "earth summit" shindig -- Nick Clegg has announced that starting in 2020, the UK will no longer use GDP data as the sole measure of economic progress.  I wasn't aware that we currently did that, and I'd be the last to deny that GDP is a pretty ropey statistic -- more of which below -- but what Clegg appears to be proposing seems downright daft.  No real surprise there, I guess.

According to Professor Clegg, GDP will give way to something known as GDP+, which will take account of the impact of economic activity on the nation's "natural wealth". Just how will that work? If a particular activity uses up non-renewable resources, will that be deducted from "GDP+"?  Or will it be even more airy-fairy than that, as the linked article from the Telegraph website rather ominously hints?  Would anyone really think, to quote the example in that article, that there's a reasonable way of deducting damage to the Chiltern Hills from the economic growth that the HS2 railway project might create?  Who gets to quantify the damage?  The local NIMBYs, who would presumably assess it as being close to infinity,  on the basis that it would be effectively irreversible?  Or the people who use the new train line, who would assess it as almost zero,  on the basis that they themselves didn't actually live in the Chilterns?

The key problem with GDP as it currently stands is that it aggregates hard-to-measure things in a way that all too readily obscures any real meaning.  If an economy produces ten apples and ten oranges, we can say that its GDP amounts to twenty pieces of fruit, but even by doing that, we've lost a level of information that may be crucial.  In aggregating production for a much more complex real world economy into a GDP figure, that becomes much more of a issue.  If you then decide to stick in all sorts of unmeasurables -- resource depletion, loss of amenity for people near train lines, erosion of mediaeval gargoyles as a result of increased air pollution, or whatever -- you come up with a measure that's both intrinsically unreliable and impossible to interpret.

None of this is meant to suggest that the existing GDP data are ideal, even for the relatively narrow purposes for which they were designed.  The ONS has reported that UK real GDP fell by 0.3% in the first quarter of this year, a number that seems wildly at odds with all of the more upbeat anecdotal evidence that has come to hand, especially the sectoral PMI surveys compiled by the good folks a Markit.  Yesterday we received a labour force report that casts even more doubt on the GDP estimate.  Between February and April, unemployment fell by 51,000,  while the number employed rose by 166,000,  with big increases in both private sector jobs and full-time employment.  That's simply not consistent with the GDP numbers, a conclusion that still holds true even if you regard employment as a lagging indicator of growth -- after all, the economy was supposedly shrinking in the final quarter of 2011 as well, so just how long is that lag supposed to be?

Meanwhile, what about the current quarter?  The sectoral PMI data have generally moved lower, which sounds bad, but today we learn that retail sales jumped 1.4% in May, a development apparently attributed in part to warm weather during the month, a happy meteorological event that I must somehow have missed.  Individual stores, especially in the consumer electronics field, are also offering some upbeat trading reports.

So what does all this tell us about GDP for the current quarter?  Frankly, who knows?  And that's why it's important not to become fixated on GDP as the "sole measure" of economic progress, to use Nick Clegg's term again. By all means let's look at better indicators -- but I don't think we should rely on Clegg and his pals to tell us what those might be.