The meaning of GDP – part two

This is my Perspective column from this week’s Fund Strategy magazine.

In the last issue of Fund Strategy (7 August) this column argued that many criticisms of GDP are disingenuous. Such detractors merely point out that it does not incorporate other measures such as inequality or happiness. Yet there are already ways of measuring these other phenomena.

Such comments are akin to criticising, say, a Ferrari for not being good at carrying heavy loads of shopping back from the local supermarket. Although the statement is true it misses the whole point of fast cars. Those who want to use their car mainly for large family shops are better off buying a Ford Mondeo or similar vehicle.

Of course many of the critics do not object to GDP so much as economic growth as a political priority. But if that is the case they should say so explicitly. They should not be hiding behind arcane arguments about economic measurement as a way of avoiding overt political debate.

However, some criticisms of GDP are more sophisticated. They point to potential weaknesses of the measure rather than simply identifying what it does not cover.

One of the most serious is that GDP does not take into account externalities. In economics these are effects on third parties that are not reflected in the cost of the product. Pollution is a classic example. When someone pays for a new car the cost does not generally take into account the damaging effects of emissions. Indeed green taxes are often justified on the grounds that they are a way of taking into account the effect of negative externalities.

Although negative externalities do exist the critics typically forget that there can be positive ones too. For example, a more mobile society, where lots of people have cars, can have many benefits. It allows people to travel more easily and further than they could do previously whether for work, domestic chores or pleasure.

Another common confusion about GDP is that it can be compared to the sales of companies. Time and again articles are written, often by campaigning organisations or naïve journalists, stating something along the lines that 51 of the world’s top economic entities are companies rather than countries. Such claims go back at least as far as a paper by Sarah Anderson and John Cavanagh from the Institute for Policy Studies, a think tank based in Washington DC, that was published in 2000. The implication is that international corporations nowadays often tower above the nation state.

But such claims are like comparing apples and kiwi fruit. This is because GDP is technically a value added measure. This might sound technical but it has important implications for how it is understood.

Take the example of car production to illustrate the point. Why Globalization Works, a 2004 study by Martin Wolf, gives the example of a Ford car that uses tyres from Bridgestone that in turns gets steel wire from Bethlehem steel. If national income statisticians simply added the sales of all three companies the steel would appear three times: it would be triple counted. Statisticians avoid this problem by instead taking the value added by each company at each stage of the process.

If this approach were applied to companies the value added would only be a fraction of sales. If countries and companies were then compared on the same basis then corporations would fall substantially down the rankings.

Even on the unadjusted figures the largest countries and even many middle-sized ones dwarf the largest corporations. On an adjusted basis it is even clearer that nation states are, on average, far larger economic entities than even the biggest companies.

Finally, let us take the preliminary estimates for GDP published by Britain’s Office for National Statistics. These are published four weeks after the end of the quarter to which it relates and account for about 44% of the total data sample.

Some critics claim that such estimates are wrong but that is a misleading way to look at it. The ONS only claims that such figures are a first stab at the figure based on limited data. It is perhaps akin to trying to guess the result of a football game on the basis of the score in the first half. As long as it is understood that the half time guesses are only predictions it should not be surprising that they sometimes turn out to be inaccurate. It would be strange to ban pundits from making such predictions or lay into them if the final result is different.

The examples of misconceptions about GDP covered in this survey are far from exhaustive. But it should be clear that the measure can be a useful economic indicator as long as its limitations are recognised.