However, the idea that inflation in a sense underestimates its "true" rate clashes with the findings of a recent working paper by Joanne Cutler of the Bank of England's External MPC Unit. In work commissioned by DeAnne Julius, she estimates a measure of "core" inflation that includes those goods and services whose prices tend to continue rising once they have started rising (or conversely, persist in falling once they have started). These elements are most important and form the core because their persistence determines the future inflation trend. If their prices are rising rapidly now, it suggests that prices will continue rising rapidly in future.

The paper reports that since 1988 this inflation measure has been on average 0.3 per cent less than the target measure, RPIX, with the gap widening recently. In February "core" inflation was just 1.2 per cent, compared with actual RPIX inflation of 1.9 per cent. While the paper does not recommend that this excess of actual inflation over "core" means interest rates should be lower, the fact that it had one of the MPC's "doves" as its sponsor led some City economists to leap to precisely that conclusion It is an over-enthusiastic leap, however. Different measures of inflation all measure slightly different things.

If the new "core" measure is systematically below the current target measure, as it seems to be, then switching to the new measure as an alternative might just imply a lower target, perhaps 2.0 per cent or 2.25 per cent rather than 2.5 per cent. It depends what figure corresponds to prices being roughly stable on average, given the well-known difficulties of measurement and the need for a bit of inflationary lubrication to make the economy run smoothly and adjust easily. But if that figure is 2.5 for RPIX inflation, it will be less than 2.5 for an inflation measure that is usually less than RPIX.As Mr Lewis points out, the reason that the new "core" measure has been lower is that it gives a high weight to clothing and footwear, a highly competitive market where prices have often been falling in recent years. We all buy clothes, whether that includes baseball caps or not, but it is not obvious that overcapacity in fashion retailing and the unusually strong impact of foreign competition on the clothing industry have important implications for monetary policy It is after all just one industry.

For the purposes of comparing demand pressures with supply pressures in the economy generally, it surely makes more sense to use the conventional price indices that weight items by what people actually spend on them, despite the cappuccino problem noted above.Ms Cutler judges her new "core" inflation measure by a different criterion, namely how well it helps forecast future inflation on the target measure. If it does this well, it could be indirectly useful for interest rate policy, which depends on what the MPC forecasts inflation will be in two years The answer is it will do moderately well in this forecast. So the "core" measure is a useful weapon to add to the Bank of England's forecasting armoury. But a new definition of inflation that is even lower than the conventional one does not mean that we have all been happier than we thought all this time.

Any measure is at best a flawed guide to the basic purpose of economic policy, which is advancing the well-being of the greatest possible number of people. However, Mr Lewis will be happy to know the price of salad cream has actually been falling.d.coyle independent.co.uk.   Bass's £810m purchase of Compass's Posthouse hotels marks the first time the leisure group has dipped into its pocket to make a major acquisition since raising £2.3bn from the sale of its brewing business to Interbrew last year.Bass has identified hotels as its core growth business and in the past nine years, it has shed 6,500 of its 7,500 pubs to focus on brands such as Inter-Continental, Holiday Inn and Crowne Plaza. Bass intends to sell 15 of the 79 Posthouse properties, while the rest will be converted to the Holiday Inn brand.