Lies, damned lies and inflation statistics

As inflation tumbled to a mind-numbing zero per cent earlier this week, the question on many people’s lips is: why is everything still getting more expensive?

The simple answer lies in the tired old joke about the statistician "who drowned in a pond that was only 1-inch deep – on average". The Retail Price Index, or more to the point, the Consumer Price Index, is no more than a measure of how fast the cost of an average basket of purchases has changed over a period of time, usually 12 months.

It's given by the simple formula ((Pi – Po)/Po)) x 100, where Po is the price of the basket a year ago, and Pi is the price of that basket now. If the average shopper spent £200 a month on groceries a year ago, and £204 on those same groceries now, then as far as most of us are concerned, the rate of inflation is two per cent.

(Not so economists, who would probably accept that definition of inflation on pragmatic grounds, but will continue to argue til the cows come home as to what inflation "really is").

Of course, that is only part of the story. One source of confusion is the way in which government happily bandies about two quite different statistical measures – Consumer Price Index (CPI) and Retail Price Index (RPI) – and the fact that these are now a long way apart. So whilst CPI is currently running at 3.2%, RPI is 0%.

Worse still, the newspapers are starting to report inflation measures that vary according to age or social class. It is possible to have too much information.

Let’s start with how that notional basket of prices gets put together. It is an average shopping basket supposed to represent what the average Brit spends in a month. As the Office for National Statistics (ONS) puts it (pdf): "a convenient way of thinking about the RPI is to imagine a very large ‘shopping basket’ full of goods and services on which people typically spend their money: from bread to ready-made meals, from the cost of a cinema seat to the price of a pint at the local pub, from a holiday in Spain to the cost of a bicycle".

Categories of spend included in the calculation range from necessities – such as food, utilities and cost of accommodation - through to what are usually regarded as discretionary purchases: entertainment, holidays and electronic goods.

The ONS carries out a massive monthly sampling process to track movements in each item and category. They also apply a weighting process to compensate for the extent to which one or other item features in the national shopping basket. As a nation, we spend more on petrol than bubble gum: so a small rise in the price of petrol will have a much larger effect on the indices than a massive rise in the cost of bubblegum.

Keeping consistency between these shopping baskets, year on year, is rather more difficult, and is managed through some rather clever statistical sleight of hand. Some items – such as videos and video recorders – can flit in and out of the price indices in the space of no more than a decade. Other items may gradually appear (or disappear).

This month (pdf) saw the addition of rosé wine, hot rotisserie chicken and internet-based DVD subscriptions, whilst wine boxes, MP3 players and rentals from DVD hire Shops were booted.

Many items included in the original RPI when it was launched in 1914, including candles, mangles, back-lacing corsets, tram fares and shirt collars, have dwindled or vanished altogether.

That index was created to protect workers from the 'temporary' economic consequences of the war, and its origins as a stark and ever-so-nannyish cost-of-living index can be seen in the fact that most of its indicators were food or food related. Alcohol was excluded altogether and tobacco received only a small weighting (share) of the total spend, well below the level it actually merited.

The nanny state, with its distaste for unhealthy living and anti-social purchases, is not a New Labour invention.

The CPI is much newer and was created at least in part as a means to create cross-European comparisons for prices as a precursor to monetary union. It differs from the RPI most in the areas of council tax and owner-occupier housing costs, and the fact that it is calculated as a geometric mean, rather than a simple arithmetic one. In theory, that should make the CPI a more accurate reflection of changes in what consumers spend and usually lower than the RPI.

The fact that it suddenly is much higher reflects the recent dramatic falls in interest rates: if you have loans or a mortgage, you are likely to be (much) better off. If you haven’t, then prices are heading upwards.

Unfortunately for Gordon Brown, although RPI is used to set many public policy decisions, such as state pensions, welfare benefits and index-linked government bonds, the measure that government has chosen to highlight as an indicator of its success over the last decade - and which has been used to target bank interest rates - is CPI. Just when he needs good news most, his favourite measure is turning out to be a dog.

And even the "good news" was less good than hoped, as it turned out that the actual inflation levels released this week were higher than experts had been predicting just one week ago.

As to why some of us are feeling poorer: none of us are average. If you had a high mortgage , large (variable rate) loans and lived on your own, the chances are that you would now be feeling relatively affluent: otherwise, the days of low to negative inflation to come may turn out to be pretty illusory.

But don't take our word for it: the ONS provides its own personal inflation calculator. So if you'd like to know what the current turmoil is doing to your purchasing power, just download the handy calculator provided here. ®

Corrections

The original version of this article got its simple formula wrong and referred to the "Cost Price Index" instead of the "Consumer Price Index". Our thanks to the many people who spotted the errors.