Everybody knows inflation is rising. What’s less appreciated, though, is that the level of consumer prices (and therefore the annual change) is not so much an objective fact as a measure which is both theoretically controversial and distant from many of our actual experiences.
The CPI is a bundle of goods and services, in which items are weighted by the spending patterns of the average household. But of course, there is huge variation around most averages. Electricity, gas, coal and oil have a weight of 3.3 per cent in the CPI. Poorer households, however, spend a bigger fraction on them. And with these costs rising 23.2 per cent year-on-year now, they face a higher inflation rate than the 5.1 per cent headline rate.
This issue of how to reduce different experiences to a single number is especially tricky for housing: how do we average housing costs to encompass people paying one-third of their income in rent and those living in mortgage-free owner-occupation? Of the CPI, 9.4 per cent is devoted to rent, but owner-occupiers’ costs don’t feature. They do appear in the ONS’s CPIH measure, but in a peculiar way. The ONS measures them by the rents tenants pay on comparable properties.
But this is controversial: if you are an owner-occupier seeing rents in your area rise, do you really think you’ve become worse off in the same way you are if your gas bill goes up? Probably not. But there’s no obviously better method. As the ONS says, measuring owner-occupiers’ costs “is one of the most contentious issues in the field of inflation measurement”.
But not the only one. If the price of one good rises sharply, we can reduce our personal inflation rate by spending less on it. One advantage of the CPI over the old retail prices index is that it better captures the tendency for consumers to do just this. What it doesn’t capture, though, is that substituting away from more expensive things is easier in some cases than others. Sure, if the price of apples rises we can buy fewer of them and more pears or bananas. But what do we do when the price of gas rises?
A further issue is the treatment of rarely-bought items. Cars have a higher weighting in the CPI than electricity and gas, because the minority who buy them at any one time spend a fortune. But this means that, for most people, an index that includes cars doesn’t measure their spending pattern and hence their inflation.
This used to mean that experienced inflation was higher than reported CPI inflation, because the latter was dragged down by falling prices of infrequently-bought goods such as TVs and computers. With car prices now rising 12 per cent year-on-year, however, ONS inflation is over stating experienced inflation for some of us.
Macroeconomists used to be well aware of these issues. Maynard Keynes wrote that the “element of vagueness” in the concept of the general price-level “makes this term very unsatisfactory for the purposes of a causal analysis”. Talk of such a thing, he wrote, was “mock precision.”
And for many purposes, such talk is unnecessary. Millions of economic decisions are made not on the basis of aggregate data but local ones: what’s happening to my or my company’s prices and costs? What’s happening to my own income? And so on.
Macroeconomic data, then, are of limited accuracy and value. And don’t even get me started on the idea of “capital”…
