In Greatly Dubious Premise I, we highlighted some of the curious inconsistencies inherent in the maths behind an economy’s gross domestic product (GDP) number – particularly what elements are included in the definition and how they are valued. Two further fundamental issues relate to technological innovation and exchange rates – how do the GDP statisticians work these into their calculations?
The question of the degree to which technological innovation can affect GDP really came to the fore in the late 1990s when it became clear that, while computers were becoming much faster and more powerful, they were not growing correspondingly more expensive. But of course technology did not begin with the invention of the silicon chip.
Take artificial light sources as an example – 500 years ago, say, humans would have had to rely on candles made out of tallow or beef fat to see in the dark; today we have LED bulbs. The things is, while GDP can easily take into account the number of candles or lightbulbs sold, it struggles to adjust either for improvements in quality or reductions in price.
In a straight comparison of tallow candles and LED bulbs, it is estimated that GDP overstates the price of light by some 900x while underestimating the quality of light by 1,600x. A similar sort of comparison of computer power and costs, meanwhile – this time over just a few decades – would be out by a factor estimated to be between one and five trillion.
The solution GDP statisticians have hit upon involves something called ‘hedonic pricing’, which is designed to capture improvements in quality and changes in price. By now, however, you will have realised that very little is simple when it comes to GDP and we will consider how hedonic pricing can affect our perceptions of the relative health of economies in Greatly Distorted Premise III.
Pretty much everything we have covered so far on GDP has been at a national level but of course it is important to be able to compare different countries’ GDP numbers on a like-for-like basis. Given the different currencies often involved, this requires some sort of market rate. Unfortunately, the market exchange rate is only based on the rate required to clear the internationally traded imports and exports, not the rate required to compare entire economies.
Comparing traded goods such as rice or grain may be a relative breeze for the GDP statisticians but what do they do about something like a haircut? Nobody trades haircuts internationally, of course, but they still need to be taken into account in GDP calculations – as does the fact the price of a trim is going to be very different if you are comparing, says, Sudan and the US.
The way the statisticians have found to account for all the pricing differentials between economies is another adjustment and this one is called ‘purchasing power parity’ (PPP). In theory, it is the solution to this problem but – and of course there is a ‘but’ – it has the effect of raising the GDP of economies with cheap, non-traded items, which are likely to be poorer countries.
This in turns makes them less likely to receive subsidised loans from institutions such as the World Bank, which is not a massive incentive to provide accurate figures for the surveys on which the PPP calculations are based. The most recent price survey took place in 2005 and covered 93% of the global population across 143 countries – of which 21, including China and India, saw falls in GDP in excess of 40%.
There could of course be any number of reasons for that but what is much less debatable is that GDP numbers are nowhere near as solid or authoritative as they might appear from a newspaper headline. As soon as you start digging away, you find a whole raft of problems – from definitions and valuations to PPP and hedonic pricing – and we will explain just why you should care in our final part of this trilogy.