Don’t get too hung up on inflation measures


Index number experts are the economic analogue of trainspotters. John explains why the Chancellor should direct attention to a broader range of monetary measures instead of redefining the inflation index.

You buy a share for 100 and a year later it is worth 1,000. That is an annual return of 900 per cent. Next year, it loses 90 per cent of its value and falls back to 100. What is the average annual yield?

There are two schools of thought. The arithmeticians say that the average of 900 per cent and minus 90 per cent is 405 per cent. Their adversaries, the geometers, think that conclusion is ridiculous. You are back where you started and your average return is zero.

Most people would agree that the geometers have the better of that argument. But suppose your portfolio consists of two shares, each purchased for 100. A year later, one is worth 1,000 and the other 10. The market has turned 200 into 1,010. The arithmeticians must be right to describe this as a 405 per cent annual return.

Arithmeticians rule the stock market indices and for a long time they also controlled inflation measures. If you spend 100 on rail tickets and 100 on television sets, and the price of your season ticket rises to 1,000 and that of your TV falls to 10, it would seem your cost of living has gone up from 200 to 1,010 – by 405 per cent, the arithmeticians would say.

But not so fast, say the geometers. If that happens, a lot of people will stay at home and watch television. If they still spend the same proportion of their income on rail fares and on television sets and do not attempt to buy the same bundle of goods, they will fare much better. The geometers’ conclusion that a tenfold rise in one price and a tenfold fall in another cancel out may be vindicated.

But can we allow central banks to say that inflation is not a problem because people can mitigate the effect of price increases by shopping around? Perhaps we need different principles to measure price increases from those we use to judge the cost of living.

Inflation indices in the US and Europe are blends of arithmetic and geometric principles. But recently the geometers have been gaining the upper hand in the Bureau of Labour Statistics and Eurostat (yes, the agency under corruption investigation). And the geometers are about to win an important victory in Britain, where the finance minister will shift his inflation target from the Retail Price Index to the (more geometric) Harmonised Index of Consumer Prices. Because geometric means give lower answers, the inflation rate will, at a stroke, fall by about 0.5 per cent per year.

The rate of inflation is not a fact but a judgment; and the range of reasonable judgments is at least as wide as the range covered by the inflation target. Today, some prices are rising, some falling; and for many goods, such as cars and pharmaceuticals, prices are going up but the suppliers claim the products are better. When inflation is 20 per cent a year, the details of its measurement are of secondary importance. At 2 per cent, they matter a lot.

The geometers claim their measures better reflect the complexities of modern life. What has happened to inflation when Harry Potter and the Goblet of Fire (paperback in the UK: £3.49 to £5.99) is superseded by Harry Potter and the Order of the Phoenix (hardback: £8.49 on Amazon but £16.99 at recommended retail price)? Or when the Dell Dimension 4600, with 256 Mb and 3.0 GHz, is replaced by the Dell Inspiron 5100, with 512 Mb and 2.8 GHz? Or when washing-up liquid comes off special offer at your favourite supermarket but a rival promotes its brand with 50 per cent more free? Answers, not to the editor but to the Office of National Statistics, which must regularly resolve these issues.

It is hard, and perhaps meaningless, to try to extract a common indicator of monetary condition from this bewildering range of information. We used a thermometer (the inflation rate) to monitor a fever (inflation) that has now subsided. It is wise to continue observation to ensure the disease does not recur, but it is foolish to attach much significance to small fluctuations, which are very difficult to interpret on close examination. The finance minister would do better to direct attention to a broader range of monetary measures than to redefine the inflation index or inflation target.

Print Friendly, PDF & Email