A Measure for Error
After a busy month I finally have a few days away for a bit of ‘R and R’, and a chance to catch up on some reading. Of course, ‘time out’, ‘switching off’ and ‘escaping the office’ are phrases a businessman does not fully comprehend, and as such my holiday reading matter is made up of the likes of LBS’s alumni magazine and Jim O’Neill’s The BRIC Road to Growth.
It was a feature in the March edition of Financial World though which got me thinking. The article by Professor Diane Coyle of the University of Manchester questions the validity of GDP. Now, this is an old debate, but I was interested in Coyle’s analysis of how GDP fails to represent the whole picture, yet continues to play a key role in decision making. I thought I would use this post to relay – to those of you who didn’t pick up this month’s magazine – the key limitations Coyle identifies in using GDP. In doing so, I will also be addressing one of my bête noires: the disproportionate prominence given to ratings systems.
A grievance with core economic statistics is perhaps an unusual stance for someone who prides themselves on being data driven. Yet I am also a man of literature and my concern is how people use “key data” without understanding the greater storyline. Indeed, I have previously expressed concern about the university and country league tables, which play a significant role in shaping opinion, although they are often subjective and miss the wider picture.
GDP was an important coefficient, but, as Coyle argues, its validity has diminished as a result of its own success. It has become an indicator of well-being and the health of an economy on which governments, institutes and companies across the world base their estimations and perceptions of a country or region. Coyle bluntly asserts “this obsession with the exact figure is nonsense”. Even though this statement is slightly sensational, I would like to echo her reminder of why GDP was needed in the first place. It came about during the Second World War to ascertain the “productive capacity of the economy” which was vital to the war effort. Thus GDP was established as a measure of economic activity, not economic well-being.
Since then the scope of GDP has expanded. What was once a 50 page booklet to configure GDP is now a tome reaching in excess of 700 pages. And as Coyle correctly points out, despite this lengthy description GDP is still subject to three key limitations: statistics, human error and interpretation.
GDP data is collated from a variety of sources such as official government records or specially commissioned surveys. The problem here is not only are the figures arguably prejudiced, but moreover they fluctuate greatly between seasons and regions. Indeed anyone in retail knows taking sales figures in February would be very different to any results from December.
2. Human error
The margin for error for the GDP figure in the UK – a place widely accepted to have ready access to high-tech and accurate measuring devises – is 2%. This is shocking considering this percentage equates to the order of magnitude of the actual headline growth.
As the GDP handbook lengthened, so the influence of certain sectors has expanded. Coyle explains how the figure is subject to pressures from industries, especially finance, which have long struggled to be recognised as a key contributor to GDP. After all, “the greater the contribution to growth any sector can claim, the greater its influence on policy”.
It is this third limitation which was particularly striking to me. It means year on year, or decade on decade comparisons of GDP are impossible as the parameters change to incorporate new influencing parties. The other issues Coyle raises are: Why does GDP include an estimation for government activity, yet the “informal domestic economy” is ignored? What about the significance of “human creativity” as reflecting growth as books get published, medicines discovered or the expansion of the Dulux paint chart? Conversely there are also negatives of growth; pollution, environmental concerns, and wealth inequality. Look at Moscow and St. Petersburg, the former has a GDP which is twice as high as the latter, but does that make the capital a pleasanter place to live?
Unfortunately, I do not have the answers to these questions. I like Coyle’s suggestion of the “dashboard” method which uses numerous indicators, but only time will tell how it is employed by those who rely on GDP. Personally, I would advocate GPD is returned to its original usage, as a measure of economic activity. There’s only one thing I know for certain, the 722 page GDP handbook will never feature on my holiday reading list!