Skip to main content

Trade Statistics

Volume 181: debated on Wednesday 21 November 1990

The text on this page has been created from Hansard archive content, it may contain typographical errors.

9.53 pm

I beg to move,

That an humble Address be presented to Her Majesty, praying that the Statistics of Trade Act 1947 (Amendment of Schedule) Order 1990 be made in the form of the draft laid before this House on 7th November.
Towards the end of the 1980s, it was widely thought——

On a point of order, Mr. Speaker. Groups .of Conservative Members all over the House are busy plotting about next week's ballot, but I am sure that you take exception to groups of Conservative Members plotting in the Chamber and taking no notice of the legislation being discussed

I doubt whether the hon. Members are plotting; I should think that they are discussing with great interest the debate that is about to begin

Further to that point of order, Mr. Speaker.

During questions to the Leader of the House last Thursday, the hon. Member for Bradford, South (Mr. Cryer) said that he believed in nuclear and other disarmament, so we cannot understand what interest the debate could have for the hon. Gentleman

I hope that all hon. Members remaining in the Chamber have an interest in the debate

It is an exciting subject, Mr. Speaker.

Towards the end of the 1980s, it was widely thought that the quality of the national accounts had deteriorated. Three criticisms were generally levelled: first, that wide discrepancies had opened up between the three measures of the gross domestic product; secondly, that successive measures of GDP and the balance of payments were subject to substantial revisions; and, thirdly, that the sector balancing items had grown unacceptably large, especially in the personal and the overseas sectors.

Those weaknesses created difficulties for economic policy makers, particularly for the then Chancellor of the Exchequer in constructing his 1988 Budget. The problems were clear to Treasury Ministers. They were also obvious to the Select Committee on the Treasury and Civil Service, which commented on the role that the poor quality of the statistics had played in recent economic policy making. The Committee recommended that the Government should undertake an investigation into the operation of the various Departments involved in the collection of national account statistics with a view to improving the reliability of those accounts.

In June 1989, the Government announced the setting up of a scrutiny committee of economic statistics. That was carried out the following year, and led to the publication in 1989 of what has come to be known as the Pickford report. The scrutiny revealed an extensive awareness of the problems with micro-economic statistics, and reiterated the three main areas of concern—discrepancies between GDP measures, revisions and large balancing items. It concluded that the problems were
"deep seated, pervasive, have multiple causes and have existed for a long time."
The scrutiny committee proposed a number of sectors where improvements were needed, or where further work was necessary, and made recommendations in three areas: first, to change the way in which macro-economic statistics are collected or compiled; secondly, to propose further work on statistical issues to identify improvements; and, thirdly, to create a statistical organization—[ Interruption.]

I am grateful to the hon. Gentleman. This is an exciting subject.

The third improvement suggested was to create a statistical organisation more likely to meet the needs of users in a cost-effective manner.

As a consequence of that third group of recommendations, we set up a review to examine the organisational issues. On 5 April 1989, the Prime Minister announced to the House the enlargement of the Central Statistical Office as a separate Government Department responsible to the Chancellor of the Exchequer. The reorganisation, on 31 July 1989, brought together the old CSO, the Department of Trade and Industry's business statistics office and the Department of Employment's responsibility for the retail prices index and the family expenditure survey. It enabled the new CSO to pay more attention to priorities for improving the national accounts.

It soon became apparent to my right hon. Friend the Chancellor of the Exchequer that even those improvements would be insufficient to solve the underlying problems quickly enough. In April 1990, he told the Treasury and Civil Service Committee that he was
"concerned about the statistical base".
We needed,
"the best level of statistics we can get for measuring precisely what is happening to the economy."
Despite the Pickford improvements, he said that he was discussing with the director of the CSO what might be done to improve the general quality of economic statistics.

Shortly after that, on 17 May, the Chancellor announced his package of further measures to improve the quality of statistics. Those were to be in three sectors—services, companies and the balance of payments. On services, there will be extensions to the quarterly turnover inquiries into the service industries and more information will be collected on external trade in services. To improve company statistics, the Chancellor announced that more quarterly information on capital expenditure, stock-building and profits would be obtained.

On the balance of payments, the Chancellor announced that, in addition to improvements on trade in services, the quarterly direct investment inquiry would be expanded. More information would be collected about United Kingdom companies' financial transactions with overseas residents. Due to the very severe problems with the balance of payments accounts, which had shown large and persistently positive balancing items over the past four or five years, he announced that the CSO would give a senior member of its staff a year to undertake a thorough review of balance of payments statistics. Those additional inquiries, including that on company profits, will be undertaken as statutory inquiries.

The powers available to the CSO to collect these statistics are embodied in the Statistics of Trade Act 1947. That Act enables "competent authorities"—certain Government Departments including the CSO—to require undertakings to provide statistical information on specified areas of their activities.

The Act includes a schedule which lists the topics on which competent authorities can demand information. Section 5 allows additions to be made to the schedule by Order in Council. Such orders have been made on two previous occasions. However, the schedule, even after these amendments, does not include profits. Therefore, I am seeking the approval of this House for a further amendment to be made.

Company profits make an important contribution to the income measure of gross domestic product. They represent about 15 per cent. of the income measure of GDP. Perhaps more importantly, they are the most volatile component of income, accounting for a large part of the quarter-to-quarter or year-to-year changes in GDP.

Statistics on company profits provide the base from which corporation tax receipts are forecast. They are also important in providing the Government with assessments of the financial position of the company sector. Finally, the statistics are used by Treasury economists in their analysis, modelling and forecasting of the economy.

The Government already compile statistics on company profits, but there are some significant weaknesses in the existing statistical systems. The CSO currently estimates the profits of industrial and commercial companies from Inland Revenue data on assessments for corporation tax. Although substantial adjustments have to be made to bring them into line with the national accounts, these statistics are thought to be soundly based. These data, however, are inevitably not available until some time after profits are earned and also relate only to whole years, not to quarters. For monitoring, forecasting and policy purposes, quarterly data are needed.

In an attempt to meet those needs, tax assessment data are supplemented by a quarterly inquiry asking a sample of companies for information on their profits. Results from those inquiries provide figures for the national accounts for the latest two to three years. The inquiry is voluntary, and many large companies exercise their right not to provide data. Despite several efforts to recruit more companies, the inquiry remains unrepresentative, and lacks a number of large company groups.

As a result, estimates derived from the present quarterly inquiry are weak, and subject to large revisions when the tax assessments data become available two or three years later. For example, the 1987 estimates of industrial and commercial company profits were revised down by £10 billion between April 1988 and the summer of 1990.

It is for these reasons that I seek approval for amending the schedule. By adding the words "profits" and "losses" the CSO will be able to collect up-to-date quarterly information from industrial and commercial companies about their profits on a statutory basis. It will be able to collect statistics from all the largest company groups in the country, not just those which agree to report voluntarily

Many of us, like the Minister, serve on the Treasury and Civil Service Committee. Often private companies are asked to provide information at their own expense and to fill in one form after another, and junior people are asked to fill them in to satisfy the people who gather statistics——

All Governments have had the problem of collecting statistics and this Government have done more to try to make them respectable and clinical than any other Government since the war. But how are we to make sure that statistics are pure? Until we have a fully computerised system—and the Government pay for it—why should private companies set their best people to collecting figures?

My hon. Friend makes a valid point. Certainly the number of forms that businesses have to fill in has been greatly reduced by this Government. We have consulted the CBI and the stock exchange on this order and we think that the information that we shall request will be readily available from companies without their having to do additional work.

Statistics from a statutory inquiry would be of much better quality, not just because of the greater size of the inquiry, but because of its better representativeness. A statutory inquiry would also produce an earlier and higher level of response.

A general advantage of statutory inquiries is that they are more equitable than voluntary ones. Companies that respond to voluntary inquiries do so from a sense of public responsibility. The costs that they incur may be small, but they are still costs that will not be incurred by their non-responding competitors

It is important that the statistics on which economic decisions are made are correct, but my hon. Friend will be aware that there are considerable cost implications for businesses. He referred to the largest companies. However, it would be illogical if the consequences of his inquiries were to be limited to large companies, particularly as, under this Government, there is a booming middle and small business sector. What worries me is that this will be extended to the small business sector and will impose even more burdens on small business men, for which there will be no compensation. What does my hon. Friend have to say about that?

My hon. Friend makes substantially the same point as that made by my hon. Friend the Member for Birmingham, Selly Oak (Mr. Beaumont-Dark). We must make sure that we do not ask for information that we do not need. However, the Government must have proper economic statistics for the management of the economy. Moreover, industrial companies like to have reasonably accurate statistics about what is going on in the economy. Those inquiries are being sent only to about 1,500 companies. We are asking each of them for seven pieces of information. All that information is, we believe, readily available in their management accounts. We do not believe that we are imposing additional costs upon them, and our view is supported by the stock exchange and the Confederation of British Industry

Will the Minister confirm that some trade associations have approached the Government for information about their area of manufacturing activity? They do not have that information. As they wish to obtain it, they recognise that the only possible body who can provide it efficiently is the Government.

There is certainly something in what the hon. Gentleman says. We receive complaints about the inadequacy of the information that is available just as often as we receive complaints about companies being asked to provide information that they would prefer not to provide.

As the House knows, the Government are very sensitive to the reporting burdens that they impose on businesses. In this case, I believe that that burden must be slightly increased. The reward of better statistics is worth it. We have already consulted the CBI and the stock exchange about the proposals. Both accept the need for accurate statistics and support the introduction of such an inquiry

If it leads to better statistics, can my hon. Friend assure me that gross domestic product figures will finally be made available in their entirety for the Isle of Wight? We have suffered for a very long time from being included in Hampshire's statistics. There is considerable prosperity and industrial growth in Hampshire. Our inclusion in those statistics has caused us considerable grief on a number of occasions, because we should like to know precisely how our own economy is doing

My hon. Friend's sentimentality is admirable, but I am afraid that the answer to his question is no.

The CBI believes that the reporting burden will not be too great as most of the required information can easily be provided from existing sources in most companies. That was confirmed by a number of large companies visited by the CSO a year ago. Any new inquiry is bound to involve some companies in extra work. I shall be discussing the details of the inquiry with the aim of minimising that extra work with my right hon. Friend the Secretary of State for Trade and Industry.

That explains why the Government have sought the addition of the words "profits" and "losses" to the schedule, but we are seeking further additions: rents, interest and investment income receipts, dividends, interest and tax payments. The reason for their inclusion is mainly related to the collection of information from financial companies.

For national accounts, there is a major difficulty in collecting profits information from financial companies. To the companies themselves, a substantial share of their profits is earned by making a turn on interest rates, or on the price of instruments traded, such as shares or bonds. For the national accounts, interest and dividend flows are viewed as transfers between institutions or people and are not therefore viewed as part of profits.

There is therefore no point in sending out a survey asking financial companies for their profits. The results would be irrelevant to the national accounts. The only way to obtain the required data is to ask companies to provide certain components of their income and expenditure accounts and for the CSO to derive profits in the appropriate terms.

That approach is adopted in existing surveys of various groups of companies. However, most of the surveys are confined to the collection of annual data. Figures for the lastest quarters are simply projections from the latest annual survey. They are subject to revision—sometimes substantial—when the next survey results come round.

For example, in August this year, estimates of financial companies' profits were revised down by £4 billion when annual data were substituted for the quarterly projections.

The further additions to the schedule will enable the CSO to block this gap in its statistical base by collecting income and expenditure information from financial companies on a quarterly basis. I do not believe that expansion of collection would add significantly to the burdens on businesses. Most of the information that is likely to be sought is readily available from most companies' accounting systems.

The phrase in the proposed amendment on services is included to make it clear that such information falls within the scope of the Act. Amendment of the schedule will enable the CSO to undertake statutory inquiries of companies to collect information on their United Kingdom profits. The information thus provided will be of significantly better quality than the present estimates. As part of the package of improvements announced by my right hon. Friend, this will be a further step in eliminating the three major problems in the national accounts. It will help to reduce the gaps between the GDP measures; it will reduce the scope for revisions; and it will lead to more coherent accounts for the company sector by reducing its balancing items, particularly for the early estimates. This will be an important contribution to the improvement in the statistical base for economic monitoring, forecasting and policy making.

I commend the motion to the House.

10.10 pm

It is ironic that the last business of the House before the censure debate should be an attempt to reform one of the principal culprits blamed for the Government's and the country's economic difficulties—the disparities and balancing items in the national accounts in recent years, to which the Economic Secretary referred.

To avoid such major mistakes in economic policy as occurred in 1988, it is necessary to have reasonably good statistics, a reasonably good model and reasonable objectives. However, none of those can be blamed for the economic policy mistakes of 1988. It is also necessary for Ministers to pay attention to the evidence, and to use the system of policy analysis intelligently.

I sent a paper to the Treasury shortly after the crash in October 1987, of which I sent a copy to the Economic Secretary this morning, reporting an exercise that I had done on the Treasury model using the Treasury's own policy optimisation programmes on the data and model available at the time. The policy adjustments required, on the Government's own priorities, to take account of the stock market crash were indeed a reduction of interest rates, but by less than 0·5 per cent. and for one year; and a reduction of income tax—but at 0·1 per cent., it was so small as to be effectively zero. The trouble was that the analysis of the situation before the stock market crash was already calling for increases in interest rates and income tax, but that was not acceptable to the Government and they were not prepared to listen. I am sure that comparable information was given to Ministers by Treasury officials.

Certainly, better characterisation of the effect of the housing market, structural reform in housing finance and consumer behaviour were needed. Changes have now been made in the Treasury model. There are now fresh problems—convergence within the exchange rate mechanism, to name but one. The Treasury has even refused to equip itself to model the problem of competing policy makers with different objectives. Whether it be between countries, within the United Kingdom or even within the Cabinet, the problem of conflicting policy makers would seem to be of some topical interest to the Treasury.

The quality of the Treasury's economic policy analysis, and, more important, the use made of it by Ministers, is not so high as to justify the cut of two thirds, from £1·5 million a year to less than £500,000 in the next academic year, in the funding of external research on economic models by the consortium of the Economic and Social Research Council, the Treasury and the Bank of England. Those funds support the work of the London Business School, the National Institute of Economic and Social Research, Patrick Minford in Liverpool, other friends of the Government and the smaller projects that are producing the new ideas for the future. That is an essential service to the Treasury and the Bank of England, from which they have benefited enormously in recent years.

On the "customer pays" principle, the Treasury and the Bank of England should have taken up the financing of their quasi-executive servicing when the ESRC found it hard to meet urgent new demands for work on the global environment and other matters.

Certainly, however, better statistics are needed, too. Household and consumer behaviour is an important matter, with which the Economic Secretary did not deal and which is not our business tonight. We are dealing with company profits and finance, and services used. Four information systems describe the same activities in the corporate sector: company internal management accounts; company reports and accounts published under the Companies Acts; tax returns to the Inland Revenue; and Government statistical returns, not to mention the returns of banks to the Bank of England under the Banking Act 1987.

It is easy to say that the different systems serve different purposes, but they interact and their incoherence reduces their efficiency and increases the cost to businesses of complying with them. It would be much simpler for companies if their management accounts giving a true and fair account of their companies' position could serve as the basis from which all the other requirements could be met. Attempts at economic and statistical analysis of companies' present reporting of their profits have fallen on stony ground. I doubt whether Government statisticians will be able to make much sense of them, with their jumble of different accounting conventions, varying from company to company. The Institute for Fiscal Studies and the London Business School have put a great deal of effort in this direction. If they are not able to make sense of the picture, I do not see how greater insight could be brought by the Central Statistical Office.

In tabling the order, the Government do not seem to have informed or consulted the Financial Reporting Council and the Accounting Standards Board. The Bank of England recently submitted a paper to the Financial Reporting Council on corporate reporting, summarising three major United Kingdom research studies. Edwards, Kay and Mayer, in their book "The Economic Analysis of Accounting Profitability", have given a convincing economic rationale for accepting value-to-the-owner rules as the basis for the valuation of assets for both economic analysis and company taxation purposes.

The Government should use their powers, not just under the Statistics of Trade Act but, probably more importantly, under section 228 of the Companies Act 1985, to specify a form and content of company accounts which will tell a coherent story about the company to managers, shareholders, investment analysts, employees, creditors, the Inland Revenue and Government statisticians and economists. It will still be necessary to set accounting standards, but the new accounting standards bodies will be given a coherent basis on which they can work. The accountancy profession has not found it easy to agree on a number of conventions, and a lead is needed from the Government in the interests of economic management, good company management, efficient capital markets and realistic wage and pricing behaviour, which needs to be based on information on the profitability of individual companies.

The claim to commercial confidentiality has always been used to resist every extension of disclosure in the interests of greater efficiency. We are talking not about personal information, where there is a basic right to privacy, but about the activities of corporate bodies. There should be a presumption in favour of disclosure. In any particular case, the onus of proof that the preservation of confidentiality is in the public interest should be on those who are arguing for it.

This makes possible a different approach to the integration of the four information systems. The basic report should be the company report and accounts produced by the company for its own purposes to satisfy its own shareholders. Accounting standards working on the foundation of the updated requirements of schedule 4 to the Companies Act should make the published report and accounts a true picture of well designed management accounts within the firm. Statistical returns to Government should be consistent with these and stated to be so in the company report, with the statistical returns subject to audit at least on sample basis.

Already, the Bank of England can call for the auditing of prudential returns required under the Banking Act 1987. If corporation tax definitions were brought within the same system, it would be possible to see much more clearly the effects of corporation tax and changes in it on company behaviour, as well as the response of corporation tax yield to changed economic circumstances. That would improve the quality of the tax base and provide a clearer climate in which companies could compete, take over and merge, and in which competition policy could be formulated and conducted.

The concept of the national accounts should not be sacrosanct. Their purpose is to facilitate the analysis and management of the economy. The apparatus for economic policy analysis has developed enormously since the system of national accounts was conceived, essentially in its present form, 50 years ago. Modern information technology makes it possible to base macro-economic policies on micro-economic behaviour, examining the aggregate of effects on actual companies or on samples of actual households, which is what the Treasury, the Department of Social Security and the Inland Revenue do today. The concept of the national accounts has now been replaced by the much wider date set, with a variety of analytical and modelling tools that can be used to ask particular questions.

Government use accounts for only a small part of the total use of business information and decision support systems. But the Government have a vital part to play in developing the statutory and operational framework for such systems, most of which will operate, in a distributed way, throughout the economy.

Opposition Members support the present order, but it is only a first and partial step towards giving business and the Government the modern information and decision support systems that they need. I gave the Economic Secretary notice of some questions with which I should like him to deal. I am grateful for the replies that he has given thus far, although I cannot say that they are very satisfactory or reassuring. Opposition Members believe that the House will have to return to these matters in a wider context.

How frequently does the Economic Secretary propose to publish the profits and incomes statistics that he will be collecting under the order, what degree of disaggregation will he offer and will he distinguish between revenue in the United Kingdom and the international revenue of companies?

Finally, let me say in the hon. Gentleman's support, on the question of the burden of the collection of statistics, that he is entirely right that, in some cases, it is more satisfactory to have a statutory basis than a voluntary one. Inquiries carried out by the Department of Trade and Industry and the Central Statistical Office provide good evidence to support that assertion. Those inquiries were first reported in British Business on 21 July 1989. Following extensive telephone interviews, an attempt was made to appraise firms' attitudes to completing statistical forms. British Business states:
"he results do not suggest that the majority of firms regard statistical form-filling as particularly burdensome. Only about a sixth of firms contributing to voluntary inquiries expressed any sort of negative attitudes in their answer to the most directly relevant question … The scrutiny team also noted that only 12 per cent. of all respondents experienced difficulties in completing forms."
A survey of attitudes to reduced statistics of trade under the new Community provisions gave an even clearer picture. Only 39 per cent. of firms said that they were willing to provide voluntary data, but 71 per cent. said that the reduction in detail was damaging to their business. Some 50 per cent. said that it had no effects, and a further 21 per cent. no significant effects, on costs. Some 66 per cent. preferred monthly returns to quarterly returns, and 71 per cent. preferred regular periodic returns to occasional returns.

Conservative Members, when speaking for small companies, often—quite understandably—reflect the views of owner-managers, who are immensely busy and do not want to be bothered, and who play a small part in the overall statistical collection system of Government. In bigger firms, and in many specialist small firms, information collected by Government is essential to the business, not on the secretarial or accounting side, but on the marketing and business planning side. If the Government and hon. Members asked them about their requirements for Government statistics, they would learn that those people would be entirely in favour of improved quality, greater accessibility and greater accuracy.

I therefore hope that the Minister will not be discouraged by some of the reactions of his colleagues and will press on with the improvements in statistics, bearing in mind that a very much more radical reform is required in the interests of both the Government and business.

10.25 pm

I note with interest, but without criticism, that the Minister tonight has the advantage of six advisers. The substance of the text of this order comprises seven lines, which I make gives 0·8571428 officials per line. That is of course an entirely useless piece of information, but it could be used for the preparation of statistics to show how many advisers Ministers have on the average night in the House. It is an example of the sort of methodology which political considerations could impose on statisticians under the politically dominated regime for the collection of statistics which we currently have.

Useless statistics do no harm, but politically massaged statistics, of which we have seen much in recent years, can do great harm because they mislead. Above all, they can mislead those who need them most, in particular those concerned with managing industry or dealing with exports.

It is clear from various reports, including the Pickford report to which the Minister referred and which was commissioned by the Cabinet Office, that the Government have had many problems with their economic statistics in recent years. Their forecasts have been seriously out and many measured indicators which should have been giving the same message have diverged. In particular, the measures of gross domestic product expenditure and GDP income have diverged.

The changes proposed in the order, in addition to placing yet another form-filling burden on industry, fail to resolve the real problem underlying the collection of Government statistics and the regime in which Government statisticians operate. I would also be interested to know from the Minister how many of the pieces of information that are now going to be asked for, in the survey of 1,500 companies to which he referred, would be available from a proper examination of the accounts that companies file at Companies house. Is adequate use made of the information provided to Companies house? I wonder whether Companies house is rigid and severe enough on those who do not file their accounts in proper time.

The Royal Statistical Society remains critical. According to Professor Peter Moore, the chairman of one of the Royal Statistical Society's working parties, there is some justification in the widely held public view
"that much statistical information is not of the quality that many users would expect and that, in some instances, it is being used subjectively in the domain of public policy."
That means it is being massaged for political reasons by Ministers.

On 28 July, The Economist, a magazine which is frequently not unfavourable to the Government, reflected that
"most international boffins reckon that Britain's statistics—once the world's best—have deteriorated more than anybody else's."
That appeared in an article headed "Numbers not worth crunching", a title which gives a clear idea of the journal's view of Government statistics.

I support several trenchant criticisms that have been made by the Royal Statistical Society. I can divide them into three headings, the first of which is autonomy. In terms of formal constitutional arrangements, the autonomy of the United Kingdom's statistical system is poorly protected by statute and practice compared with most other developed countries. The order does nothing to improve the situation. In many countries, the statistical agency has total control over publications, with an embargo, generally, on anyone having access until the statistical results are publicly available. That embargo includes the Government and the Minister who is notionally responsible for the statistical agency. That should be the arrangement in the United Kingdom, but the United Kingdom entry in the United Nations handbook makes it clear that the Government statistical service lacks control over publications. It states:
"Decisions on publication of statistical results are the responsibility of the (policy) Department, but the Head of the GSS will be consulted on the more important issues."
Thus it is the case that the Government statistical service is not proof against charges of bad news diversion because its operational environment denies it satisfactory control over the nature and timing of publications. The order does nothing to address that problem.

The major functions of data collection, processing, primary analysis and publication should be ring-fenced and centralised away from ministerial interference. The head of the statistical service should manage the operations and the staff of the service, including his appointees in ministerial Departments. A statistics Act, such as exists in most western countries, is needed to guarantee the autonomy and constitutional posit ion of United Kingdom official statistics.

The second heading is methodology—again something which is not improved in any way by the order. Questions of methodology, definition, scope and timing of publications should be the responsibility of the head of the statistical service. That would strengthen the perceived independence of Government statisticians and the line of professional responsibility to the head of the service on technical matters. Professional identity is weakened in this country because, unlike most other countries, we currently have a decentralised Government statistical service in which most statisticians are bound to have dual loyalty with the service on the one hand, but with line management responsibility to the policy departments— that is to say, the departments in which most of them are embedded—on the other.

It is important that a research unit be set up to strengthen evaluation and methodological research. That would do more than the proposals in the order to improve the reliability of and the credence given to Government statistics.

The third heading is accountability. To put it another way, to whom can the statistician run if he finds that the statistics that he has prepared are being massaged by policy departments for political reasons? At present, the only public mechanism available to the head of the Government statistical service in response to a ministerial decision which might compromise the integrity of the service appears to be a threat of resignation, whereas Government accountants, if they are asked to do something that conflicts with their professional principles, are entitled to ask the Minister for a written instruction which is brought to the notice of the Public Accounts Committee. What is really needed is a national statistical commission to be appointed to safeguard the objectivity, integrity, timeliness and scope of United Kingdom official statistics. It should report annually to Parliament via the sponsoring Minister so that Parliament can be assured that it is free of ministerial interference at any time and at any level.

This is an important order because such great use is made of Government statistics. They are used for the purpose of business, industry and the City in particular. They are also frequently used in the House for the justification of political argument. Unfortunately, they are not always accurate. There are ways of making them much more accurate, but the order fails in any way to address the ways in which that may be achieved.

10.34 pm

The hon. Member for Motherwell, South (Dr. Bray) referred to the cut in the budget for work on the Treasury model. That decision was made by the Economic and Social Research Council, not by the Treasury. We are committed to paying 20 per cent. of that budget and would have been willing to pay 20 per cent. of a higher budget. That was the council's choice and priority.

Did the ESRC ask the Government's chief economic adviser whether the Treasury would increase its contribution to make good the cuts in the funding so that the ESRC could see its way to providing what was essentially a service to the Treasury?

I do not know the answer to that question, but I will let the hon. Gentleman have an answer. My understanding is that the ESRC made the decision and the Treasury would have been willing to pay its 20 per cent. share, as it has done in the past, of whatever the ESRC chose to spend.

It would be nice to have one set of company accounts rather than separate ones for corporate reporting, tax returns and statistics, but that would not be possible or practical. They serve different purposes and there is no such thing as a correct set of company accounts. Many different assumptions can be made, and for tax purposes there are particular rules on depreciation or what expenses are allowable for tax purposes. Therefore, it would never be possible to have the same annual accounts for shareholders as for the Inland Revenue.

On statistics, there are major differences—for instance, on the inclusion of overseas activities of United Kingdom groups which have to be disaggregated for the national accounts, the treatment of depreciation which is properly deducted from company accounts but not from national accounts, and the treatment of interest and dividend flows which are different for national accounting purposes.

The book to which the hon. Gentleman referred lays out a plan—largely for inflation accounting, as I understand it—but accounting standards bodies have not been able to agree on such matters for the past 20 years and it is highly unlikely that it would be right or proper for the Government to attempt to impose a particular standard of accounting and reporting which they felt would be suitable for the many different purposes. That is not a practical proposition.

Has the Economic Secretary asked either the Financial Reporting Council or the Accounting Standards Committee for their attitude to increased legislation in this matter?

As I have said, accountants and accounting standards boards cannot agree. The United Kingdom Accounting Standards Committee does not agree on all sorts of things. Its biggest difficulty was on inflation accounting. Company accounts should be clear and explain through their notes how they arrive at the figures presented. It is perfectly easy for any reasonably numerate person to interpret such accounts in a way which gives access to a good deal of information about a company. But to put the accounts of leasing and manufacturing companies, overseas traders and banks into some straitjacket which applies to all of them is impractical. The information will be taken quarterly. The foreign and United Kingdom profits will be disaggregated and we are asking for seven items of information, all of which will be made available.

I am grateful to the hon. Gentleman for what he said about the burden of the collection of statistics. We have reduced by half the number of forms that businesses have to fill in—from 800,000 to 400,000. I regret that this order and the Chancellor's initiative will result in a few more, but we are confining it to the 1,500 largest companies and we expect to get from them the vast bulk of the information that we need because that is where most of the corporate profits are.

The hon. and learned Member for Montgomery (Mr. Carlile) made a lot of the Central Statistical Office not being independent and of politically massaged and managed statistics. I work a lot with officials of the CSO, they are a thoroughly professional group of people and what he said is pretty offensive to them. There are absolutely no grounds for saying it. The Royal Statistical Society's report said that it found absolutely no evidence of what the hon. and learned Gentleman suggests. If he suggests that the statistics are manipulated politically, he is completely wrong

The hon. and learned Gentleman must let me finish. He made a serious allegation and I intend to convince him that he is completely wrong. There is absolutely no way in which Treasury or any other Ministers could interfere with statistics. We have no basis for doing so and we have no grounds on which to do so. They are collected by and analysed by professionals and the publications are put together by professionals.

The dates on which the vast majority of statistics are published are laid down well in advance. If any Treasury or other Minister instructed a delay, it would rapidly become apparent. It has never been the case. The hon. and learned Gentleman should be extremely careful about making allegations of that kind, which cannot be substantiated

The criticisms that I made were directed not at the professionalism of the statisticians but at the regime in which they are forced to operate. Moreover, is the Minister aware that every word that I said in the debate is supported by the representations of a working party composed of senior officials of the Royal Statistical Society? Why is the Minister not prepared to pay attention to that distinguished body's representations on these matters? I can show him documentary evidence of their support, if need be

The Royal Statistical Society said that it found no evidence whatever that

"individual government statisticians operate to anything less than the best professional standards".

No, I should like to respond to that point. The hon. and learned Member for Montgomery suggests that the statisticians allow themselves to be manipulated by politicians like me

It is what he suggests because the statistics could be manipulated only by politicians such as myself telling statisticians to do something other than what they would professionally want to do

I am grateful to the Minister for giving way again. I do not wish to prolong the argument, but does he not appreciate that the regime is such that if Ministers—I make no allegations against the Minister—wish to manipulate the statistics, there is no guarantee of the statisticians' independence? That is the complaint which the Royal Statistical Society rightly makes. Why are the Government not prepared to give the same guarantee of statisticians' independence as exists in most western countries, as he well knows? Why can we not have a Statistics Act to guarantee their independence?

The hon. and learned Gentleman now makes an entirely different allegation. He says that the system is such that it is not clear that the statistical service is independent. He said before that Ministers manipulated statistics

I am glad that the hon. and learned Gentleman is no longer making that allegation. I was coming to the system.

If the Minister had read on from the quotation that he gave, he would have read:

"but we are not satisfied that the organizational framework in which they"—
that is Government statisticians—
"work offers the best protection against undue pressure."
Earlier in the report the society said:
"Integrity has not only to exist, but to be visible."
It says further:
"the lack of a standing mechanism within which government statisticians can have a constructive dialogue with producers and users and meet criticism is itself a comment on the existing arrangements."
Ministers have frequently discontinued series of statistics and surveys which they found politically embarrassing. For instance, not only unemployment but the distribution of income and wealth have been scandalously treated by a Government who think that they do not matter

but there was a strong call for us to do so. It is not necessary to continue to collect information by product categories as we did previously. Articles are published annually on the distribution of wealth. The most recent one was published last week. Articles on the distribution of income are also published annually. If the hon. and learned Gentleman suggests that the regime under which the Government statistical service is managed does not exclude any possibility of interference, that is different from his original allegation

No, I shall not give way again. I have given way to the hon. and learned Gentleman a lot. He made two separate allegations——

Of course I was listening.

We believe that it is not necessary to change the system. We do not want a centralised statistical service because it is right to collect many statistics in the Departments to which that information relates. That is the way in which it has been done for a long time under Governments of all persuasions.

One of the purposes of the collection of statistics on, for example, transport is to enable transport policy makers and Ministers to make decisions. The utility of that to the national accounts is obviously important. The business statistics office, the family expenditure survey and retail prices responsibilities were transferred to the Central Statistical Office, but it is not our intention to transfer any more. We do not think that a centralised service is the right way to proceed. It would be inefficient and expensive and it would not be as useful to Departments as the present system. [ Interruption.] It is independent as it is. The collection, publication and analysis of the statistics are carried out by professionals. I have repeatedly said that to the hon. and learned Member for Montgomery.

The main, though not the only, purpose of the collection of the statistics is to help Government policy makers and forecasters. Everything that the statistical service collects is published as soon as it is ready. The delay in publication of the series in which most people are interested is between 24 and 48 hours after the information becomes available. There is no lengthy delay and the dates on which such series are to be published are set out well in advance. If there were any attempt to delay publication of the retail prices index or balance of payments figures, everyone would know. We do not think that it would be right for the Government to lose control of the process. The Government should be able to set the priorities and decide where resources will be allocated.

I was asked about independence. Many things would happen if a Minister attempted to interfere. The permanent secretary who is head of the Central Statistical Office has access to me, to the Chancellor and, at the end of the day, to the Prime Minister. There is no question but that in the long run remedies are available. As I have said, there is no way in which a Minister such as myself or the Chancellor could interfere in one of these series. What would one ask to be done? Could someone be asked to knock half a point off the retail prices index? That is simply not a practical proposition and there is no basis on which to do it. We do not have an alternative source of information on which we could or would seek to argue a separate case.

The points made by the Royal Statistical Society are misconceived. That is not the way we run our statistical service, nor is it the way we want to run it. Statistics are not produced for the statisticians in the Royal Statistical Society. They are produced mainly for the Government, to enable us to operate our policies. I am happy with things the way they are and with the way matters are organised. It is right that matters should stay as they are.

I take it that the hon. Gentleman has the leave of the House to speak again.

I did not refer in my speech to the report of the Royal Statistical Society, which was not directly germane to the debate, but I am glad that the hon. and learned Member for Montgomery (Mr. Carlile) referred to it. The Government's reply to that report and to technical questions about the development of statistics illustrated their incompetence in economic management and were quite appalling. The complacency with which they approach government and the needs of business lies close to the root of the Government's problems and the economic failures of the past two years. The House will have to return to these matters at greater length another time.

Question put and agreed to.


That an humble Address be presented to Her Majesty, praying that the Statistics of Trade Act 1947 (Amendment of Schedule) Order 1990 be made in the form of the draft laid before this House on 7th November.