Previous SectionIndexHome Page


Inflation Indices and Statistical Errors

Motion made, and Question proposed, That this House do now adjourn.--[Mr. McNulty.]

1.28 pm

Mr. Edward Davey (Kingston and Surbiton): This is the first time that I have spoken while you are in the Chair, Madam Deputy Speaker. May I welcome you to the Chair and to your new appointment?

I am grateful for the opportunity to hold the debate. Many colleagues will have seen its title and considered it to be one of the least interesting subjects, but I believe it relates to some of the most important challenges that face this country's senior policy makers. The United Kingdom and other countries may be measuring inflation and related economic statistics incorrectly, which may be undermining our economy's prospects for growth. Simple measurement errors could have genuine effects on the fortunes of families and businesses in every constituency throughout the country.

As a result of those errors, interest rates could be too high, the investment decisions of international capital might be biased against Britain, and jobs might have been unnecessarily lost. That is a bizarre but alarming claim. If experts outside the House who present that argument are right, and countries such as the United States are benefiting from incomparable international statistics while the United Kingdom is suffering, the Government must not delay resolving the matter.

Given the interesting developments in Florida, I am not sure whether we should wait for an American recount, even of economic statistics.

In case listeners or readers think that I have taken leave of my senses by making such outlandish claims, I immediately refer to a report by the Office for National Statistics, which was published this year with the snappy title, "Review of Short-Term Output Indicators". A small part of the analysis used in that ONS report relates directly to the claims I am making. Using a United States-style inflation measure, the ONS estimated growth in the United Kingdom's manufacturing output to be a massive 6 per cent. higher since 1995 than the recorded figures show.

That one reference to a highly credible source should suggest that that is not a scare story. Many other respectable analysts point to the problem, so those are serious and alarming claims. My understanding is that many senior people in the Treasury and the Bank of England share my grave concerns, but, to my knowledge, this is the first time that the issue has been raised on the Floor of the House.

Before I explain the background to my argument, I reassure those statisticians from the ONS and elsewhere who might be interested in this debate that I am not attempting to take a wide swipe at economic statistics in this country--far from it. The ONS report reviewed the quality of short-term output indicators in the UK and was generally very positive. We owe a debt of gratitude to the many hard-working and talented statisticians in the ONS, Departments and the Library. The statistics that we receive from them are of high quality.

Previous concerns include the divergence between the various measures of gross domestic product in the late 1980s and, more recently, the average earnings data series, but they were the exceptions that prove the general rule.

17 Nov 2000 : Column 1232

Despite the title of the Adjournment debate, I am not making a wholesale criticism of the public sector's statistical service. In fact, my argument rests not on criticism of it or the economic statistics it produces, but on how to achieve comparable international statistics for some key economic variables. The debate is an attempt to point out that the lack of uniformity might be unnecessarily handicapping UK plc. It would be absurd, and a concern that statisticians would surely share with us, if mere statistics could be said to be undermining economic policy making or performance.

The central question, which involves the background to the debate, is about how a country should measure changes in prices of goods and services, when those goods and services are changing in quality. That technical question about inflation measurement has major policy ramifications. The problem of adjusting inflation measures for quality changes has always existed. Consumers in 1990 were buying different items of different quality from those bought by consumers in 1890, let alone in 1790.

However, some people now argue that the problem has become much worse. Major improvements in computer technology and in communications technology make it even more difficult to compare the price of what consumers buy in one year with the price of what they buy the next, especially because purchases of information and communications technology products have taken up a larger share of consumer expenditure. That is especially significant because such products have become a major intermediate input cost to other services and goods that we buy, whose quality is also being continually enhanced by those technologies. It is argued that quality-adjustment problems for measuring inflation have become far more significant in recent years.

If every country had reacted to those problems in exactly the same way, there would be no difficulty for anyone. After all, price changes are only signals of relative changes in supply and demand, and if measurements of aggregate price changes for each country dealt with new developments in the same way, arguably nothing would have been lost. However, different countries have reacted in different ways to the measurement challenges of ICT. In America, the way in which inflation is measured has been altered recently, mainly as a result of recommendations from the 1996 Boskin commission. Key changes involve new ways to adjust for quality improvements in high-technology goods.

So-called "hedonic" pricing methods have been adopted. They strip out the effects of quality improvements in, for example, home computers, cars, clothes and televisions in new ways. Those methods break down a product into its key features, such as the memory and speed of a computer, and assign prices to those features rather than to the whole product. Thus, by controlling for quality improvements using such new techniques, hedonic inflation measures tend to show significant price reductions for technology goods, lowering overall recorded inflation for countries using those methodologies. Therefore, inflation in the United States may seem lower relative to inflation in the UK and most other European countries, but some or all of the difference may be explained simply by differences in how the statistics are compiled and how different countries adjust for quality changes.

17 Nov 2000 : Column 1233

Some people may ask, "So what? Why should we worry if the Americans want to record inflation differently from us, so that it looks lower than ours?" The next stage of the argument is where it gets interesting.

Inflation figures are not just some theoretical measure existing in a void, published in a press release and reproduced in economic tables, gradually to be consigned to the top or bottom shelf of the Library. Inflation figures affect real things. They impact on wage negotiations, for example, and decisions on pension rises, as the Government know only too well. They impact on the decisions of those who set interest rates--the Monetary Policy Committee--so they impact on mortgages paid by families throughout the country. Presumably, if our inflation is measured higher than it could have been measured, mortgage rates--real things--are higher than they need to be.

Again, inflation rates of different countries affect the decisions of international capital about where institutional investors wish to place their cash. If the United States' inflation measure is flattering its economy and ours is under-selling Britain's, are we missing out on our share of international investment funds? Is the wider European economy and, dare I say it, the euro being undermined by over-optimistic statistical signals from the United States?

Those questions are probably impossible to answer, so the Minister can relax. I ask her rhetorically to illustrate the enormity of the issues that might be involved by different countries treating economic statistics differently. For the record and to save the Minister the trouble, I am not arguing that Britain or other European countries now measure inflation incorrectly. I am arguing that perhaps the United States is wrong, and perhaps all the countries are measuring inflation wrongly. It is a difficult concept. There is clearly more than one way to measure inflation, but the key policy issue is: why are we measuring inflation differently and could that be damaging the UK's economic prospects, at least relative to America's, just because of statistics? Perhaps we should be worrying.

I say "perhaps" for reasons of modesty. It is a complex issue and I am no statistician. I would not wish the Minister to think that I am so confident of my case that I know that it is having damaging effects. It may not be. I simply raise the real possibility that it could be, in which case the Government need to act.

Therefore, I look for two possible answers from the Minister. Either she will say that I am nuts, that I have been bamboozled by the latest economic theory and that we should not worry because Great George street and Threadneedle street have it all under control; or she will say that there is an issue of substance behind some of the points that I make and that other people have made outside the House, that the Treasury is concerned about it and is giving it serious thought.

The only indication that we have had so far of what the Treasury thinks about the matter is contained in box 1.2 and annex A of a report, published with the pre-Budget report, entitled "Productivity in the UK: the Evidence and the Government's Approach." I welcome that report. I share many of the Government's concerns about the UK's productivity performance. I do not wish Ministers to think that, because I am questioning official statistics, especially their relation to America's, I seek to undermine that

17 Nov 2000 : Column 1234

productivity thesis and the stated goal that we need to do more to improve our economy's productivity. We can argue over the size of the productivity gap between America and the UK, but the fact that there is one seems relatively uncontentious.

I am concerned about the way in which the Treasury deals with the measurement issue in that publication. It is as if, because the measurement problem works against the paper's main thesis, it is stuck in the annex and the narrative tries to down play the significance of the measurement problem. I am not sure whether that is a good way in which to examine the issue.

It may be true, as the annex suggests, that the simplistic application of hedonic pricing measurements to the United Kingdom can overstate the differences that result. The paper quotes a recent OECD report, which suggests that other factors offset the impact of the simplistic application of hedonic measurements, reducing the apparent biased output and productivity figures. Although those academic charges and counter-charges are interesting, some senior people are worried about the issue. That is presumably why, in annex A, the Treasury writes:


All that gobbledegook means that the Treasury is trying to push ahead and solve the problem, but they are not quite saying so.

Yesterday, when Sir John Kingman was giving evidence to the Select Committee on the Treasury, I asked him what he intended to do. I hope that I do not paraphrase his words too much, but he replied that he took the issue seriously and would look into it. I apologise to the Minister for being unable to stay to hear her evidence to the Committee and for not asking her some of today's questions yesterday. I hope that she told the Committee that she shares Sir John's concerns.

Although it is good that people are now waking up to the issue, I remain troubled by the timetables for the research and debate needed to put it right--there does not appear to be one. When will we get to the bottom of this? When will formal discussion papers be published? When will the changes take place? The issue matters because, if the argument is correct, it is having very real effects, now. I want the Minister to reassure me that the Treasury is making it a top priority and that the Treasury will give every possible support--if necessary, extra resources--to the National Statistician and the Statistics Commission, to resolve the problem.

It is instructive to remember how fast people moved when problems were perceived with the average earnings index. The Treasury did not wait very long then. I should like a similar fast-track resolution to take place now and the Chancellor to make this into a public issue to give it the profile that it needs. Just by giving the matter the oxygen of publicity, some of the potentially damaging effects--from appearances, not based on reality--could thereby be corrected.

I should like the Minister to go further than simply talking about timetables and how seriously the Treasury are taking the matter. I want her to tell the House that she will draw this debate to the Chancellor's personal attention and to suggest that he focus the attention of his European counterparts on the issue. He and his

17 Nov 2000 : Column 1235

international colleagues should demand that the OECD and Eurostat do not drag their feet on the issue, but get a move on.

We can all understand that major changes to vital series of economic statistics must be undertaken carefully, without political pressure, and, in this case, with maximum international agreement. There is no dispute about that, yet I am seriously worried that the Americans have--wittingly or unwittingly--already stolen a major march on the United Kingdom and Europe by changing their data. Perceptions engendering expectations can have real effects, even--perhaps especially--in sophisticated economies. It is perhaps ironic that the measurement of information technology has led to misinformation. Let us consider whether the US's over-optimistic data have contributed to the golden halo effect around the wider US economy.

Is that effect and the data at least a partial explanation of the enormous surge in the US stock market, as the dot.com mania appeared to be based on solid macroeconomic data? Was it then also an underlying cause of the subsequent fall in values, as investors realised that many of the individual firms that they had backed during that gold-rush dot.com period were actually poor prospects?

Let me stress that I do not doubt that the extraordinary performance of the US economy is built on more than mere statistical manipulation, but it is possible that some of the hype has been. It could have caused some of the froth that has sucked world investment funds out of Europe into the US. Let us face it, Alan Greenspan--the chair of the Federal Reserve--has made a virtue of not worrying about US inflation because of the USA's good productivity data. So the Americans have not waited too long to make those supporting corrections to their data with respect to the new economy. The improved perception that they have given has brought real benefits, drawing in capital and magnifying America's first mover advantages in ICT investment.

This is not the only area of economic statistics relating to new technology and e-commerce that the Government should be concerned about: I know that the Office for National Statistics is concerned about others. A member of the Monetary Policy Committee, Dr. Wadhwani, is worried about other measures. There are large differences in accounting for software investment within ICT between the United States and the United Kingdom and Europe, which could have a knock-on effect on the way in which we measure productivity. In a recent speech, Dr. Wadhwani highlighted concerns about the different measurements of capital stock. It is tempting to develop those arguments, but I hope to do so on another occasion when I try to catch your eye, Madam Deputy Speaker, so I shall not try your patience much longer.

This is a serious problem. Some of the articles and discussions in the technical press may have overestimated its effect, and in taking up some of their points perhaps I have done so, too. I am willing to debate that point. I hope that they and I have overestimated it, but some real issues need to be addressed. I do not think that the problem has had the publicity that it deserves, and I hope that the Minister will reassure me that the Government will try to sort it out.

17 Nov 2000 : Column 1236

1.46 pm


Next Section

IndexHome Page