We are richer than you think

Peter Kellner finds serious faults in the official figures for the British economy but warns that, if they are revised, the poor (as usual) will be the losers

Which European country has an inflation rate of just 1 per cent? Whose economy has an underlying growth rate, still improving, of more than 3 per cent a year? Who are better off than the French and Italians?

Welcome to Britain.

To describe our country in this way is to invite derision. It contradicts the conventional story: that we are the poorest of Europe's big four countries, that our economy is incapable of growing by much more than 2 per cent a year, except in short bursts, that we are in or on the edge of recession, and that it took tough measures last year to pull inflation down to the annual 2.5 per cent target.

There are, however, reasons for believing that the conventional story is wrong - and that it is wrong because the information that shapes our perceptions of Britain's economy is defective.

Consider, for example, prostitution, drug-dealing and illegal gaming, which between them make a substantial contribution to Britain's economy. Official calculations of the size of the economy derive from such things as tax returns, VAT records, payroll data and company records. Illegal activities, involving cash-only transactions hidden from the Inland Revenue and Customs and Excise, do not show up.

The European Union now says they should be counted. Its argument - echoing an earlier United Nations ruling - is that national accounts should strive to measure all transactions involving mutual consent. Their legality or illegality is irrelevant. Nor is it relevant that illegal activities are (or might be deemed to be) "bads" rather than "goods", diminishing rather than enhancing the well-being of society. As a recent article in Economic Trends, the monthly journal of the Office for National Statistics (ONS), puts it, "the national accounts do not measure welfare; they measure economic activity".

The article reckons that conventional figures underestimate consumer spending by up to £12.5 billion, or 2.6 per cent. Illegal drugs (£9.9 billion) is by far the largest component, followed by prostitution (£1.2 billion), illegal gambling (£800 million) and the sale of stolen goods (£700 million). These estimates are hedged with qualifications and are cautious. And they do not include the sale of illegally copied videos, the growing amount of alcohol and tobacco smuggled across the Channel, bribery and fraud.

It is quite possible that once these things are all included, Britain will turn out to be as prosperous as France and Italy, which the conventional statistics say have overtaken us during the past 20 years. Some British statisticians believe that the Italians edged past us by incorporating bold estimates of their black economy. On a like-with-like calculation, Italy may not have overtaken Britain at all.

But it is not just in the tawdry world of illegality that economic statistics prove inadequate. Like many people I pay a subscription to an Internet provider. That subscription is captured by the national accounts as an economic activity. What, though, about my use of that subscription? As more and more people acquire e-mail, I send and receive more messages this way. Yet all that the national accounts record is the cost of my telephone connection to the Internet provider. Sending the same letter by post first class to ten people costs around £3 (postage plus paper and envelopes). Sending the same e-mail to ten people costs 5p. As far as the national accounts are concerned, economic activity has been reduced by £2.95. Yet as far as the purpose of my action is concerned - communicating with ten people - the same thing has been achieved more quickly. My personal output is up; but Britain's economic output is down.

The problem lies with the term "economic activity". It covers transactions in which money changes hands. The underlying assumption about measurements of national income is that this is the same as prosperity. At one level this is obviously a nonsense. Our well-being is affected by a range of non-economic factors - health, friendship, pollution, crime and so on. But "economic activity" does provide a useful proxy for measuring changes in well-being, provided that the boundary between economic and non- economic activities remains constant.

Information technology is moving that boundary. By shifting a growing amount of communication from "economic activity" to "non-economic activity", it leads to an undermeasurement of economic growth. It is the opposite to what happens when a developing economy moves from subsistence agriculture, where people grow their own food, to a cash society in which food is bought and sold. As food production shifts from a non-economic to an economic activity, the economy is likely to record spuriously high rates of growth.

Alan Greenspan, chairman of the US Federal Reserve, has observed that statisticians have failed to keep pace with the changing structure of the econom, and "the progressive substitution of intangible services for physical output". This does not affect only "unmeasured" activity, such as e-mail messages, but the treatment of investment. When manufacturing companies buy a new machine, their purchase counts as investment and is added to GDP. When a new technology company buys computer software, this counts as a production cost, not as a boost to national income.

Modern technology causes growth to be undermeasured in another way. It often fails to take account of quality improvements. Television sets are an obvious example. The price of a basic colour set has remained remarkably stable - around £150 ever since colour television was first launched more than 30 years ago. But their quality has improved vastly. They provide better pictures, they are more reliable, they last longer, they consume less power and most also provide teletext. Yet as far as economic output is concerned, a television is a television is a television.

Or consider computers. The machine on which I am writing this article cost roughly the same as my last computer, but it has twice the processing power and three times the memory.

The statistical tools we have for measuring growth are fine for measuring the production of standard, unchanging products, such as a loaf of sliced white bread or a bus fare from Westminster to the Old Bailey, but start to become rather shaky as soon as the products themselves change. With a fairly simple product such as a light bulb, we may agree that, since it will last twice as long as a light bulb of 50 years ago, statisticians should count it as "worth" two light bulbs of yesteryear. But they don't. And the reason is that the task of measuring changes in products is normally far harder. Consider my computer again. Its whizzier technology makes only a modest difference to my everyday life. It saves me time on some things (switching the machine on, opening files, using e-mail and accessing Internet pages), but not on others (composing an article takes just as long). Changes in quality affect different consumers in different ways. We can be certain that official statistics under-record the growth of the economy, but nobody knows - and perhaps nobody can know - by how much.

Just as growth is under-recorded, inflation is over-recorded. If a £150 television today is considered to be twice as good as a £150 television was 30 years ago, then the number of "units of television worth" has doubled per set sold. By the same argument, the price of each "unit" has halved. Something similar applies every time a new generation of products turns out to be more reliable and longer-lasting than its predecessor, but at the same price. Governments around the world have found it hard to incorporate such improvements in their inflation figures.

In the US, concern about the quality of inflation statistics led to a Senate-commissioned inquiry chaired by Professor Michael Boskin of Stanford University. Its report concluded that the US's consumer price index overstated the true inflation rate by at least 0.8 percentage points, and possibly as much as 1.6 points. If the higher figure were right for Britain, our current inflation rate would be hovering around 1 per cent a year, not 2.5 per cent.

Even 1 per cent may be an overstatement, for three reasons. First, the arrival of bar codes and modern checkout technology has allowed supermarkets and other shops to adopt more versatile pricing policies, such as loyalty cards and "buy one, get one free" offers. Yet the retail prices index (RPI) ignores such price- discounting. It counts only the full unit price.

Second, new products are not included in the RPI until they are well established and widely used. Computers did not show up until last February; nor did mobile phones and call charges. So the rapid price falls in the previous decade - which commonly take place with new products - did not show up in the official figures.

Third, the system that Britain's ONS uses to combine different price rises into a single, composite figure for the RPI has a built-in bias. It uses an arithmetic instead of a geometric system to weight different figures. To illustrate what this means, suppose that the RPI consisted solely of apples and bananas. Say that both cost £1 a kilo today. Now suppose that in one year's time, apples are down to 50p a kilo, while bananas are up to £2 a kilo. Suppose that a year later both are back to £1 a kilo.

You might think that the RPI would be the same in two years' time as it is today. Under Britain's system, it would not. The RPI, if it started at 100, has moved to 125 in the first year and 156 in the second. This assumes shoppers divide their spending equally between apples and bananas. So they will buy more apples and fewer bananas in the first year and are therefore deemed to lose more from the doubling of apple prices in the second year than they gain from the halving in banana prices.

Under the geometric-mean method, the RPI would indeed remain at 100. And that is the standard European Union method of measuring inflation. If EU rules (which also involve a different treatment of housing) were to be used here, our inflation rate would be around 1.4 per cent instead of 2.5 per cent.

How much does any of this matter? Were we to change the way we measure our output and inflation, we would change merely the numbers, not reality. The numbers, however, are important. They affect the way we think of our society and its problems. If the numbers are wrong, our diagnosis is likely to be faulty and the remedies mistaken.

There are also more practical consequences. For example, the Bank of England might not have raised interest rates last year if it had had more realistic inflation figures to hand.

Another consequence is potentially more explosive. The inflation rate has a direct affect on the taxes we pay and the welfare benefits that millions receive. State pensions, social security benefit rates and tax allowances normally rise in line with the RPI. If the inflation rate were reduced by one percentage point each year, the government would save around £1.5 billion in the first year, £3 billion in the second year, £4.5 billion in the third year and so on.

I make this point not to advocate it but to warn that it may be in the Treasury's mind. Last week, the ONS announced that it is reviewing how it measures inflation. I should be astonished if papers have not been prepared in the Inland Revenue and the Department of Social Security to show the savings they would make from switching measuring techniques.

Those who lost out would almost certainly be the poor. Yet one of the most important by-products of any rigorous attempt to describe our economy more accurately would be to show that the gap between rich and poor has grown even faster than the official figures suggest. Who is it that takes most advantage of the tumbling price of new products, or the arrival of the latest television and video technologies, or high street loyalty cards or supermarket "multibuy" offers? (Which income group, for that matter, spends most on prostitutes and cocaine?) If the government changes the measure of inflation used to calculate welfare benefits, then it must employ one that is directly relevant to the spending patterns of the people who receive those benefits.

The truth is that Britain's businesses, legal and illegal, are creating more wealth than our official statistics would have us believe. But it would not reflect well on Labour's priorities if the revised data that demonstrated this point were also to be used, as I fear some Treasury officials may be tempted to use them, to screw the poor.

Post this article to

• Digg
• del.icio.us
• newsvine
• NowPublic
• Shoutwire
• Reddit

We want to encourage people to comment on our content and to exchange views with other readers and hope this will be done on a courteous basis. However, if you encounter posts which are offensive please let us know by using the 'report this comment' facility or by emailing comments@newstatesman.co.uk and we will take swift action where necessary.