Skip to main content

Census

Volume 889: debated on Tuesday 25 March 1975

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

11.0 p.m.

I beg to move,

That items 7, 8, 9(d)(e)(f)(g)(h)(i)(j), 10, 11(a)(b)(c)(d), 14, 15, 17(a)(b), 18(vi), 19, 20, 21(a)(b)(c) in Schedule 2 and items 6, 8. 10(c)(d)(e)(f)(g)(h)(i), 11(vi), 12, 13, 14(a)(b)(c) (d)(e), 16 and 17 in Schedule 4 to the Order in Council entitled the Census Order 1975, a draft of which was laid before this House on 20th February, be approved.

Mr. Deputy Speaker has selected the amendment to this motion in the name of the hon. Member for Berwick-upon-Tweed (Mr. Beith), to leave out from "That to "to" and to insert:

"item 17(a) in Schedule 2".
The Chair understands that it would be for the convenience of the House if the motion which has been moved and the motion
That the draft of the Order in Council, entitled the Census Order 1975, which was laid before this House on 20th February, be not submitted to Her Majesty.
were to be debated together. I should, however, remind the House that it will not be possible for me to put the Question on the latter motion after 11.30 p.m.

The draft order providing for a census of the population of Great Britain to be taken in 1976 has been laid before the House under the Census Act 1920. It is subject to negative resolution, but some of the items of information which it is proposed to include in the census returns are not specifically authorised in the Act and require approval by affirmative resolution of both Houses. The motion covers the items concerned, which are printed in bold type in Schedules 2 and 4 to the draft order.

Since there is a Prayer against the order, we have the opportunity to debate the order in full, and I will briefly explain its provisions. The House will remember the controversy which surrounded the 1971 census. At that time it was felt in some quarters, I believe, that the content of the census and the arrangements for it had not been adequately explained to Parliament and had not been adequately debated. We have tried to overcome this problem by ensuring that as far as possible there can be no such feeling on this occasion, and the Government published on 20th February a White Paper on the 1976 census of population outlining the reasons for taking a midterm census, the topics it would cover and how it would operate. I hope this will help the House. The White Paper was an innovation—no previous census had been explained in this way—and I hope the House will agree that its publication has been useful in setting out our proposals and also in providing a basis of information for this debate.

Because the White Paper has been published I will not describe all the details of the order, but I shall be happy to check up any individual points which hon. Members may raise.

The annual expenditure of central Government and local government now represents over 45 per cent. of the gross national product. Another factor is the speed at which our society is changing. Populations of local areas can change nowadays, in numbers and characteristics, with great rapidity. More than one person in 10 changes his address every year, and this makes it important that the information available to us is as up to date as possible. Moreover, the new local and health authorities set up as a result of the major changes in local government and the administration of the health service require up-to-date information for their activities. Yet without a census in 1976 we and they would be working as far ahead as 1983 or so, when the first results of a 1981 census would be expected, on information dating back to 1971. Clearly, this would be unsatisfactory. We have a right to expect that the planners, whose activities touch our lives in so many respects, will have a better base for their work than out-of-date statistics.

It will be seen from Article 4 of the draft order that the census is to be a full count of the whole population. Because of the need for accurate information on the size of local populations and the increasing demand for statistics relating to quite small areas, a sample census as was taken in 1966 would not give us the full information which we need. But we realise that filling in forms can be time-consuming and burdensome for many people, and we have tried to keep the number of questions which each form-filler will have to answer to a minimum. This is not easy when information is urgently needed on so many different subjects, but the Office of Population Censuses and Surveys has devised and tested a sampling system which reduces the load which each person questioned must bear. This sampling system will be applied to the returns for all the 50 million or so people in England and Wales. In Scotland, with a total population of 5¼ million people, the small sizes of the populations of many local areas make this sampling system inappropriate. A single set of questions differing from each of the 10 sets in England and Wales will, therefore, be asked throughout Scotland, and to reduce the burden on the public some of the topics covered in England and Wales will be omitted in Scotland. My hon. Friend the Under-Secretary of State for Scotland will say more about the Scottish arrangements should he catch Mr. Speaker's eye later in the debate.

In 1971 a number of people asked whether the form for a private household could be filled in by the husband and wife as joint heads of the household, and we have recognised this in our draft order. Members of private households who do not wish to furnish their particulars to the head of the household will under Article 5(5) be able to obtain their own form which they can return direct to the enumerator. I can also mention that envelopes will be available in non-private establishments for those who want to seal their returns before handing them in to the management, although this is not provided for in the order.

The House will rightly attach importance to the arrangements to be made to safeguard confidentiality once the information is in the possession of the census organisation. By "census organisation" I mean the two offices which conduct the census—the Office of Population Censuses and Surveys and the General Register Office for Scotland, including the census takers who will be specially recruited for the purpose. In assigning the enumerators to districts, the local census officers will be instructed to follow the principle, as far as possible, of keeping them away from areas where they already know or are known to the inhabitants through their business or personal contacts. But there can never be a complete guarantee that an enumerator will not know or be known by any householder in his district, especially if it is in one of the more remote rural areas.

Where an enumerator and a person he is enumerating are none the less known to one another, the enumerator will be instructed to offer an envelope in which the return can be sealed before it is handed to the enumerator. In this way the return will bypass the enumerator as the envelope will not be opened by him. There will also be a procedure for sending in a form by post. The great majority of householders find the traditional method of "doorstep" enumeration perfectly acceptable and many—especially old people—welcome the help enumerators can give them with any difficulties they may have in completing the forms. All the alternatives to the normal method of handing back forms to the enumerators who check for completeness on the doorstep will add to the complexity of the operation and increase the time taken and the cost.

The Minister will agree that it would be helpful to clarify the situations in which return of the form by post will be permitted. These were described in the White Paper as "exceptional circumstances". The hon. Gentleman has not used that phrase tonight.

It is open to anyone to do this. The problem, as was explained in the White Paper, "Security of the Census," published in July 1973 by the Conservative Government, is that it adds considerably to the cost of enumeration.

Members of the "field force" will be liable to the penalties laid down in the Census Act 1920 if they divulge any census information. These penalties include imprisonment for up to two years. Staff at the Census Offices will be subject also to the provisions of the Official Secrets Act 1911. Strict physical security arrangements will be made for the transit of the documents to the Census Offices and at the Offices themselves. The Registrars General have taken account of the recommendations of representatives of the British Computer Society made after the 1971 census for increasing the protection of the computer records. The physical security of the OPCS computer installation has been reinforced and a new computer operating system with more built-in safeguards introduced. Finally, care will be taken to ensure that the statistical tables produced from the census do not reveal information about identifiable individuals or households. I have dwelt on this because many Members are rightly worried about these aspects. I hope they will consider that these safeguards are substantial.

I turn now to the amendment proposed by the hon. Member for Berwick-upon-Tweed (Mr. Beith). This seeks to delete from the order all the affirmative resolution items in Scotland and all of them except the usual address one year ago in England and Wales. Thus, in effect, the amendment would exclude from the census all except the subjects which were specifically mentioned in the schedule to the Census Act 1920. That is a list of items that were of particular importance to Government at that time. But the information needed by Government changes with time and the effect of the amendment now would be to exclude from the census a number of topics, all of which are important and many of which are related to items that the amendment does not seek to exclude. In selecting the subjects for inclusion in the census the Government have subjected all suggestions to rigorous scrutiny and have rejected any that were not of first importance to central or local government in carrying out their jobs today. I believe that to do this is far better for today's conditions than to refer back to conditions in the 1920s.

In the early-day motion on this subject the hon. Member for Berwick-upon-Tweed and other hon. Members propose that items excluded from the census be covered by a voluntary social survey or a supplement to the census form to which no penalty for failing to answer would be applied. A social survey unconnected with the census would not be a satisfactory way of collecting such information. Practically all the information in the census is interrelated—for instance, educational qualifications are classified by the occupations and industries in which their holders work—so that unless the survey's results were to be related to those of the census most of the census questions would have to be repeated in an unconnected survey. The suggestion for a voluntary section in an otherwise compulsory census form would be likely to cause great confusion and probably affect the quality of response even to the compulsory questions. That is the advice I have been given, and I have studied it carefully. We in Britain are not alone in rejecting the voluntary element of a census. Experience has led most other countries to the same view. However, it is possible to relegate some questions which might otherwise be included in the census to voluntary surveys linked to the census.

However, the distinguishing mark of voluntary surveys is that some people do not answer them and subjects can be covered in such surveys only if the level of non-response does not invalidate the results. In the voluntary pre-test of the census which was carried out last year, a response of some 80 per cent. was achieved. This is a high level indeed, and we greatly appreciate the co-operation given by the public. But in many cases it is just not possible to base policies on information where replies from two out of 10 people are missing.

This proportion far exceeds the size of the groups about which it is most important to get information, for instance, the households without all basic amenities in their dwellings and the immigrant communities. Moreover, we could never be certain that the people who did give information were representative of the people who did not. Indeed, one might suspect that they would not be. I hope that this point will be borne as carefully in mind by hon. Members who have taken a different view.

Thus I am forced to conclude that we are right to retain the element of compulsion. It is the only way of avoiding a situation in which the replies from the people who do answer the questions are not wasted by the failure of others to reply, and of ensuring that the statistics provide a sure base for policy.

The Government, however, have in mind four voluntary surveys linked to the census, two to be conducted throughout Great Britain, and the other two to be conducted in Scotland only. They will be carried out by the Census Offices. The proposed surveys are described in the recent White Paper on the Census, and I shall be happy to answer questions on them.

Understandably there has been criticism of the time it has taken to produce all the results of the last census. A very great volume of material has been produced from the 1971 census already, nearly three times what was produced from any of its predecessors. On the other hand, I acknowledge that several detailed sets of tables remain outstanding on particular topics, such as transport to work and migration. They are part of the main publication programme, and we are doing our best to ensure that they will be laid before the House by the middle of this year—just over four years after the census was taken. [Interruption.] Hon. Members may laugh but it is a formidable task to assess these statistics.

The Government accept that it is important to improve the speed at which the census results are produced. We plan to complete all the main tabulations from the 1976 census within three years. To achieve this we shall need to keep an eye on the volume of analyses that we agree to undertake so that we do not jeopardise the early production of much-needed information by diverting resources to less important statistics.

There are two main stages in the production of census statistics: firstly, the information on the forms has to be converted by clerks into a format which the computer can handle. To speed up this part of the operation we shall decentralise it so that we can recruit a larger work force spread between a number of regional offices. Secondly, there is the computer processing; a larger computer is being installed at the Office of Population Censuses and Surveys with the capacity to handle the data more quickly. As in 1971 this will do the main tabulation work for the whole of Britain; a new computer in the General Register Office for Scotland will do editing work on the Scottish data and some of the Scottish tabulations.

The House will wish to know how much the proposed census will cost. We have estimated the total cost at 1974 Public Expenditure Survey prices as £21 million for the Census of Great Britain. This does not include the cost of the follow-up surveys designed to extend the information provided by the census. Of the £21 million, £17½ million is for the Census of England and Wales and £3½ million for Scotland. In spite of the larger number of staff we shall be employing at the peak period and the enhanced computer resources, the cost in real terms is unlikely to differ substantially from the cost of the 1971 census. These are difficult decisions, particularly at a time when public expenditure is under great scrutiny. The costs are borne by many Government Departments. [HON. MEMBERS: "The taxpayer."] Of course it is borne by the taxpayer, but the taxpayer is also interested in how the Government spend the money.

This is a serious issue. Hon. Members will wish to make points of criticism at this stage rather than a few weeks or a few days before the census. The purpose of the debate is to try to ensure that the census is conducted in a way that is as far as possible free from controversy on the actual day of the census. It is in that spirit that I am taking part in the debate, so that we can sort out now rather than later any problems that exist.

I am sure that the House will agree that if the Government and the local authorities are to provide the wide range of services that the community expects of them, and to do so efficiently, they need a correspondingly wide range of information on which to base their decisions. It is the function of the census to provide this detailed information for the whole country and in a consistent manner for areas within it.

I shall be happy to deal with detailed points about the conduct of the 1971 census and under-enumeration as I reply to the debate. The success of the census is absolutely dependent on public understanding and on the public's certain knowledge that the information will be protected.

I hope that the House will join me in acknowledging the quite outstanding record of the Census Offices over more than 100 years since they were established in maintaining the high standard of confidentiality of the personal information provided by the public. The information collected in the proposed census and in the follow-up surveys I have described will be given the same high degree of protection.

I hope that hon. Members will agree that our proposals in the draft order give the country the type of census it needs in the middle 1970s.

11.17 p.m.

This is a very complicated matter. May I first say how much we welcome the White Paper, which has been a great help to us in understanding the order.

Ever since the first census in 1801 the great national head count—which is what it is--has aroused anxiety and controversy. I am sure that the Minister of State will not be too surprised if I say straight away that we are very sceptical about much of what he is proposing.

First, we have serious doubts based on the delays in presenting the previous census material. So much of this information is useless if it is delayed too long. The Minister said that there had been difficulties with the 1971 census, but I remind him that we still await the general report from the 1966 census, and volumes from the 1971 census promised for 1973 have still not appeared. According to paragraph 32 of the White Paper, material will be available within three years, but on the previous track record I very much doubt that.

Secondly, we have doubts about the form proposed. Some of the sampling of the 1966 census, which was a 10 per cent. sample, proved faulty. The Minister should be more convincing about the 10 different variants proposed this time. Will they produce similar sampling errors? Would we have done better to have another sampling process but on a better basis than the 1966 sample? Will the Minister say a little more about that?

Thirdly, there is the vast cost that is involved. I understand that the 1966 census cost £3½ million. The 1971 census was estimated at £10 million but the figure is already beyond that. The Minister now says that the next census will cost £21 million. My guess is that we shall find that figure equally wrong. Can he justify that vast expense when fewer questions are to be asked. Why are fewer questions to be asked if the questions were all so essential last time? Will there be savings in the spending of various Departments as a result of all this information? If there are to be savings perhaps the Minister will give us some examples.

Before we support the Minister I think we are entitled to ask him whether he has gone into this matter at all. Will he give us a departmental breakdown of how much will be saved? Which of his colleagues has he consulted? A number of Departments are affected and we need to know the answers to the questions that I have raised.

This census is likely to be the most complicated ever to administer. Will the staff be available? Personally I doubt whether it will be available. Is there not a real danger that it has been made so complicated that by the law of diminishing returns the errors will increase to an extent that will make much of the information unreliable? I have a number of authorities that suggest that that is likely to be the case.

Let us consider the use which is to be made of the material that is gathered? Will it justify the huge cost that is involved? I was not reassured when I read an article in Statistical News by the Deputy Director of the Census Office. He wrote:
"I see a real need to find out more about the use to which the census data are put."
Has the Minister seen that need? Has he investigated it adequately? Sweden, the Netherlands and West Germany avoid many of our census problems by having population registers. In Canada the authorities run monthly sample surveys for the same purpose. In Britain we have the General Household Survey. Why should we not greatly extend that survey, as has been done in the United States and in West Germany? Again, I must ask the Minister whether he has gone into these matters. Could we not, for example, immensely reduce the costs as well as increase the accuracy of the information?

Why are the definitions constantly changed? It is now impossible to compare even such ordinary subjects as lavatories, habitable rooms and kitchens between one census and another. In 1961 kitchens were counted as rooms only if meals were regularly eaten in them. In 1966 all kitchens counted as rooms. In 1971 they counted only if they were more than 6 ft. wide.

The OPCS appears to regard the census operation as akin to taking a snapshot of the British population on census day without having any regard to the past. We cannot compare various social classes because some occupations have been moved from one class to another. There will be no overall figures for the British Isles for cookers, sinks, hot water and school qualifications because certain ques- tions are not to be asked in Scotland. Why do we have to speak, read and write Welsh but only speak Gaelic? Why should penalties vary for non-return? Why should most fines be £25 when after the 1971 census eight Belfast priests were fined only £7? A man at Halesowen who said that the census was an obnoxious and bossy imposition was fined £15. These are serious questions.

Then there is the whole question of confidentiality. This is perhaps the most serious matter of all. I understand that many of the enumerators turned out to be Government or local government officials and —Inland Revenue officials, Customs officials, rent and welfare officials and immigration officials, many of them and—this is the most serious aspect—working in their home districts.

What has the Minister to say about that? I thought that he brushed aside too lightly the question of confidentiality. When I was in Sweden recently, the question of the confidentiality of medical records was being discussed, and the worry was that there could not be interchange of records between one region and another because of the lack of confidentiality of computer records.

Is it correct that it would cost an additional £1½ million to overcome this and that only if this money is spent can it be overcome? The Minister should be open with us about this.

I think that it will be clear that we on this side of the House have the gravest doubts about this census in the way that it is set up and about whether it will justify its immense cost. If we have this census, what prospect is there that we can avoid the next? Cannot we examine alternative ways of collecting some of the same information? Without much more convincing assurances, I doubt whether I or my hon. Friends will be able to support the statutory instrument.

11.26 p.m.

I welcome the proposal for a census, although my hon. Friend the Member for Birmingham, Perry Barr (Mr. Rooker) and I thought it only sensible that the House should be able to debate all the questions which are to be asked in the census together rather than merely debating the new questions.

There is an urgent need for a reexamination of the philosophy underlying the hold of a census. The 1976 Census of Population White Paper states:
"The object of the census is to provide statistics about the whole population and groups within it, so that government and local authorities can do their job efficiently."
I question whether that should be the sole purpose of holding a census and whether, in terms of cost-effectiveness, it could be shown that holding a census considerably improved the efficiency of central or local government.

Increasingly Parliament is considering more sophisticated and complicated legislation covering all aspects of social and economic life, and it is vital that that legislation shall be informed by an accurate knowledge of the social and economic state of the nation. If it is not, it would be possible for the House to pass a measure of social reform designed to correct a social abuse which no longer existed. That is a danger.

In addition, there is outside the House an enormous number of people engaged in doing research into the social and economic characteristics of society, trying to analyse changes in society, and it is vital that we give them accurate statistical information on which to base their research because their analysis of our society is important if we as legislators are to understand the rôle that we should play in attempting to modify society through the process of law. Their inquiries would emphasise what was modifiable via the process of law.

I am disturbed to learn that the Government, in their consultations, called a meeting of census users—I welcome the calling of such a meeting—of users who put forward some of the information which they would seek to obtain from a census. The response of the Government officials was "We called this meeting, but the census is not for you, so that it does not matter what suggestions you make." That is far too negative a view. We should see the census, not simply, or even mainly, as a vehicle for enabling Governments to improve their efficiency, although I hope that that would be one side effect, but as a major source of statistical information on which an analysis and understanding of our society can be based. That is essential if we are to pass good laws and if satisfactory action is to be taken at local authority as well as national level.

Further, there has been far too great a delay in producing the results of the census. This makes almost a mockery of some of the statements in the White Paper. Paragraph 11 headed "Housing questions" states:
"The census figures on housing are used, with projections of future populations and households, in deciding the programmes for house building, the new towns and overspill. Up-to-date information is important for these immensely expensive programmes."
If we have to wait for four years for the results of the census, including information on housing, how can the information be up to date? It is a contradiction in terms.

If we have to wait for even three years for the results of the 1976 census, we shall already be into the future which the White Paper talks about. We all know that we would be greeted with derision if in local government on a housing matter we argued on the basis of census figures which were three years out of date.

The Government should carefully consider whether it is more important to have comprehensive figures for the whole nation on a wide aspect of social and economic life or to obtain information rapidly. I believe that the census proposed for 1976 falls between two stools. It is claimed that it will produce information rapidly which can be useful to policy makers. Yet it has to be so comprehensive that it is likely to be several years before figures are produced, and this will dramatically reduce its usefulness. Therefore, if we want information quickly it would be better to concentrate on sample censuses.

On the other hand, if we want fully comprehensive information which will be of value for long-term research and policy making, there is a case for a full census and a full analysis of the results. I do not believe that it is possible easily to reconcile the two philosophies of having information which is quickly usable by decision makers and comprehensivity within the same data gathering exercise.

The hon. Gentleman is obviously a greater enthusiast for statistics than I am. Regarding the 1971 census, will he tell the House of one thing on which he changed his mind substantially because of the extra information which was given to him?

I believe that, although efforts should be made to produce the data as speedily as possible, it is clearly better to obtain the data at five-year intervals than at 10-year intervals. However, that is not to say that I believe that intervals of five years are the answer. That is why I argue that there is a serious case for examining the possibility of using sample censuses. I believe that the information to be obtained from censuses at 10-year intervals will be mainly for the benefit of historians and not for the benefit of policy makers and will be of relatively little use even to students of history when the basis on which the questions are asked is changed.

I would now like to turn to some detailed point on the order. If comprehensive information is to be sought on the question of places of residence, the House and the nation should be concerned to know the number of residences people have. How common has second ownership become? For example, how many people have holiday cottages? How many people have caravans which they use as alternative homes? We should be interested in this, because it is a major social trend in many Western European countries, and we, as legislators, will be faced with the consequences of the trend. Surely a census designed to provide statistical information of long-term trends should contain questions of that type. Such a question must be of major interest to many hon. Members and hon. Friends who represent constituencies which have not only many caravans but many holiday cottages.

It would appear that the information on occupation and employment is seriously inadequate. It is arguable that one of the major problems of our economy is that it has not changed rapidly enough. Occupations have not changed rapidly enough: industries have not been run down rapidly enough, and new ones have not been created rapidly enough.

What are we to learn from this census? There is sample information of the kind of work done for one day in 1976. In this respect the framing of this census is worse than that of 1971, which at least attempted to find out what work someone was doing a year before. If we are to get a true picture of job changes and occupation changes we need far more detailed analyses. We need questions which themselves contain a time dimension. We need to ask what employment the respondent had been engaged in for the last five years; what was the location of his employer; and what was the nature of his employment. That would enable us to see what amount of transition there had been in terms of occupations and places of employment. The census does not provide that information.

In Schedule 3 of the Census Order there are 10 sets of questions which are to be asked merely on a sample basis. I wonder whether this is an accurate sample frame, in the first place, if so many questions are asked merely on a sample basis—hopefully, on a 10 per cent. sample basis. It would have been better to have had a 10 per cent. sample census in the first place. If we want detailed and comprehensive information these questions should be answered by every household.

Again, we have a dual philosophy—the desire for comprehensive statistics, on the one hand, and the desire for sample information which is easily obtainable, on the other. I believe the proposals in Schedule 3 try to meet both criteria and fail to meet either.

I welcome the order, because I believe that information that is five or seven years old is better than information which is 10 years old; that as society is changing so rapidly, to try to base our analyses of that society, at the end of a census period, on information which is 10 years old is nonsense.

I ask the Government to think seriously about the possibility of obtaining census information on a sample basis—information that is obtainable at yearly intervals and can be made available to a wide range of users. I want the Government to be far more willing to consult users of the census about the information that they need to conduct their researches fully.

I should like to see a totally separate exercise designed to obtain comprehensive data covering a full census, which could be carried out quite reasonably at 10-yearly intervals. I think that this order, welcome as it is as an improvement on the previous situation, tries to combine two philosophies—the need for sample information and the desire to obtain comprehensive statistics—and I think that to this extent it is defective.

11.40 p.m

I beg to move, as an amendment to the proposed motion, to leave out from 'That' to 'to' and insert:

'item 17(a) in Schedule 2'.
I am grateful to the Minister of State for taking care to deal with the specific points raised not just in my amendment but in the early day motion associated with it. He will be aware of the procedural complications that make it necessary for us to have an amendment, an early day motion and a prayer to obtain a satisfactory debate on this issue. I am also grateful that the Minister took the point that I am seeking to persuade him of the desirability of treating a large section of the census on a voluntary basis.

We are discussing a proposal which has the twin disadvantage of costing £21 million—perhaps even £25 million by the time it is completed—and of making substantial incursions into an individual's liberty and privacy. It is against that background that we should judge whether it is necessary for the purpose of planning Government expenditure. Given that the Government are committed to some sort of census, I hope that in this exercise tonight we can make fundamental changes in the philosophy of the census and persuade them, through the amendment, by rejection of the order or by other means, that their whole approach to the census should be changed.

The White Paper on the census was published on the same day as the order, which does not leave much scope for any assumption that the Government will change their minds as a result of any comments. It was said on page 3 of that document:
"The first job of the census is to count the population."
I do not quarrel with that statement or its implications. It seems reasonable that the State should obtain basic information about population and citizens on pain of penalty. It should be able to establish how many people there are, where they live, and whether they are male or female. It is understandable that the simple basic facts should he obtainable on pain of penalty. But when these penalties are extended to a wider range of questions, the matter becomes objectionable.

I understand the arguments of Government Departments that they require information for the purposes of planning. I know how far those demands can go and how much more extensive than the present census any proposals might be in relation to the requirements of some Government Departments and other bodies. But there are already enough points of conflict between citizens and State without threatening individuals with fines or imprisonment because they do not want to say whether they speak Gaelic, write Welsh, did well at school, have a fixed bath, or walk to work.

There are two kinds of question I would criticise when they are compulsory. There are those questions which are acceptable in themselves but which should not be compulusory. That applies to most of the matters I have already mentioned. Most of us do not object to being asked those things in principle, but would object to being fined if we are not willing to answer them. Secondly, there are those questions which give rise to serious misgivings and which perhaps should not appear at all. Looking back to 1971, we recall questions about date of entry, relationship to head of household, and so on. Those questions caused real fears —fears that people might be made the basis of a "round-up" of immigrants, fears of past shadows hanging over people's lives that might be brought out into the open, fears among people with adopted children when the fact of adoption had not been disclosed locally. For those fears to be exacerbated by fines is unacceptable.

The Government have dropped these questions but have retained others, such as parents' place of birth and place of residence five years ago. What value can we place on this insistence when the questions which have now been dropped were regarded as indispensible. When my noble Friend Lord Avebury raised these matters in 1971, he was told that those questions must be included. They have now been dropped. Therefore, what value can we attach to similar assertions of indispensibility about questions in the present census?

There are a number of objections to the question about the parents' place of birth. In the first place, it does not tell one very much about the problems of immigrant communities or about the problem of language, and contributes to the notion that anybody with a parent born abroad is a problem.

Does the hon. Gentleman feel that the fact that this information may not be available at a later stage may give rise to all sorts of statements which, with the possession of such information, could be disproved? Providing the information is kept secret, as it is under the 100 years' rule, it can be of considerable assistance in countering propaganda which may otherwise do great damage in future.

The figures to which the hon. Gentleman referred are not about immigration as such but about specific characteristics, particularly the parents' country of birth. The Home Office collects figures on immigration. The Home Office knows how many people come into this country in a given year through the immigration procedures and the collection of those statistics. Indeed, the question to which I refer is strongly opposed by organisations concerned particularly with the welfare of immigrants who fear that these figures will be used in the wrong way. The Joint Council for the Welfare of Immigrants, the Standing Conference of Asian Organisations in the United Kingdom and the National Co-ordinating Committee of Cypriots in Britain have made objections of this kind. I quote the Standing Conference of Asian Organisations, which says:

"We have witnessed, in the past, these types of figures are being maliciously used to incite hatred against new Commonwealth born residents and their descendants, which is hardly favourable to harmonious community relations."

Does the hon. Gentleman know that this question was specifically asked for by the Community Relations Commission and the Race Relations Board, which are far more representative of opinion than are the bodies to which he referred? Do the Liberals believe that we should try to assist immigrant communities who come to this country and are suffering problems? If so, how are we to do that, in terms of allocation of resources, if we do not know where they are in this country?

That is absolute rubbish, and the hon. Gentleman knows it.

I am aware that the question was suggested by the bodies to which the hon. Gentleman referred. That illustrates a certain lack of co-ordination in the immigrant community between the bodies which formally deal with the Home Office and those at the grass roots which are in more direct contact with the immigrant community. It is disturbing that there is complete disagreement between the two bodies to which the hon. Gentleman referred and those which I have quoted.

The issue of language might have been the subject of a question if the purpose was to direct assistance in educational programmes, and so on, towards particular immigrant problems. The question is not distinguished between particular types of immigrants and problems; it is a general question, which gives rise to fears without giving the information which serves the purpose to which the hon. Gentleman referred. We could spend a great deal of time on that point if more time were available.

If the hon. Gentleman attaches such importance to this question, why is it not being asked in Scotland? The whole of Scotland is excluded from this supposedly vital question.

This question should be either removed or put on a voluntary basis.

The purpose of the amendment is to deny the Government the authority that they need to put on the form all but one of the questions which cannot be included by negative resolution. Behind it is a constructive proposal which would be even better served if the Prayer were capable of being approved. The proposal is that the basic questions about population—Questions 1 to 5—should be on the first sheet of the census form and he subject to penalty, but that the remaining questions should not be subject to penalty. They could be detachable and returnable separately, perhaps being married with the rest of the information in the computer.

There is every reason to suppose that very few people would refuse to complete the latter part of the form—perhaps not many more than the 434 who were prosecuted in 1971. The Government have begun to accept the principle of the voluntary census in the follow-up and sample surveys. Why not go the whole way? In any case, much of the census is being done on a 10 per cent. basis, so that we are not dealing with total figures for many of the questions.

There are further grounds for concern regarding confidentiality which must not go unexpressed. In particular, the Government's rejection of so many of the recommendations of the British Computer Society which had been invited to review the census following the suggestion by my right hon. Friend the Member for Devon, North (Mr. Thorpe). Indeed, in the White Paper on Security, the Registrar-General rejects many of the society's recommendations. He rejects the independent monitoring of the census.

The argument of the Government on this point—that they are waiting for the consideration of the Younger Report to be completed—is extraordinary. That report was published in 1972. We are still waiting for the completion of the Government's consideration of it. In another place, Lord Harris promised that we would have the results of the consideration of the Younger Report before Christmas. He apologised after Christmas and said that he hoped we might have it by Easter. It is now almost Easter and we have still not received the view of the Government on the Younger Committee's proposals. In any case, that is a red herring. It is not beyond the wit of the Government to make the commitment to independent monitoring of this census, even if they have not been able to complete their consideration of the Younger Committee report. The two issues can be separated, at least for the time being.

I should have thought that the Government could have made up their mind on the monitoring of the census without worrying about the Younger Committee, the results of the consideration of which we all await but which need not be wrapped up with this matter.

The rejection of the independent reception of citizens' complaints is a matter for serious concern. It is likely to pose problems during the census. Members of Parliament will receive many complaints, because there will be no other way for citizens to say "Do I have to answer this question?", "Can I send this form in a sealed envelope?", and "What are the rules?".

There are other matters for concern, such as the sale of information in small blocks, from which purchasers might be able to trace information about individuals. Then there is the problem of the enumeration staff. At the last census the enumerators had a difficult job. They were not well paid. Many of them did not know what they were in for. They found themselves with a difficult task, in a far more controversial form than they expected. I do not wish to attack them. Many of them had very poor facilities for doing their job. The census supervisor in Northumberland, for example, worked from his home and could be contacted with complaints and problems only when he was at home for meals or ready to go to bed. The working arrangements did not assist the enumerators.

Enumerators should not work in areas where they are known. We need something firmer than we have heard so far from the Government. Enumerators should not work in areas where they are professionally engaged as local government officers, as civilian employees of the police, or in any other way in which the information might be relevant to their work. We need stronger proposals on that point.

Since the Minister is unaware of the problems of immigrants —which is painfully obvious, because of the way he has dealt with the point—does not my hon. Friend agree that in some parts of the North of England, where there are huge ghettos, with hundreds of immigrants living in the area. it might not be a bad idea if immigrants were used as enumerators? Do the Government have any plans to do that?

I think that my hon. Friend's suggestion has a great deal to commend it. I hope that it has been noted by the Government Front Bench.

The constricted time available for the debate illustrates how unsatisfactory is this procedure. We need a Bill rather than an order to deal with matters of this kind. Indeed, the Census Act 1920 is due for substantial review.

It should go on record that the Government do not seem to have learned the lessons of 1971. The bureaucratic assumption is still there—as the National Council for Civil Liberties puts it—that people cannot be trusted to do anything properly unless they are forced.

The assumption still exists that demands for information should override the citizen's right of confidentiality. There is still a failure to recognise the genuine misgivings of those most qualified to understand the problems of confidentiality of computer records.

On those and other grounds, I am not willing to support the Government on this matter.

This debate finishes at 12.30 a.m. There are at least half a dozen hon. Members who wish to speak.

On a point of order. You have just said, Mr. Deputy Speaker, that this debate will finish at 12.30 a.m. May I respectfully draw your attention to what I think is the fact that it is within the discretion of the Chair to decide that the matter has not been sufficiently considered, and very seriously to put to you with great respect that it is obvious from the course of the debate so far that it is virtually impossible for the Government satisfactorily to deal with the anxieties voiced in all parts of the House? I am not in any way seeking to interfere, of course, Sir, with your discretion, but only most respectfully to bring this point to your attention.

11.56 p.m.

I am grateful to the House for its indulgence in allowing me to intervene at this stage. I shall be as brief as possible.

In moving the motion my hon. Friend referred to the Scottish arrangements, and indicated where, in the main, they differed from those in England and Wales. It may be useful if I amplify what he said.

The Government's aim in spreading many of the census topics across 10 different sample forms in England and Wales has been to reduce the burden on the individual householder as compared with the 1971 census, while at the same time obtaining sample information on a wide range of topics. Unfortunately, this extensive sampling system is not appropriate in Scottish conditions, so we do not propose to make use of it. We are dealing in Scotland with a total population of about the same size as each of the 10 samples in England and Wales and if we in turn sampled on the same scale each of our sample populations would be correspondingly smaller. Figures derived from samples are subject to sampling error. This is acceptable where the number in the sample is large, but it can introduce too great a range of uncertainty into figures derived from small samples.

Yet in Scotland we want detailed and accurate statistical information about local government areas some of which have very small populations. Examples are the islands areas—Orkney, Zetland, Western Isles—with populations of between 17,000 and 30,000 and whole districts with populations of no more than 10,000. If we are to have accurate figures of the total populations in such areas, or of groups within the population with particular characteristics and needs, such as young children or pensioners, we must have full enumeration.

This is not the first time that Scotland has been different in a census. In the 1966 mid-term census, which was a 10 per cent. sample census, several areas in Scotland were fully enumerated—Roxburgh, Sutherland, Zetland, Lewis and Harris and parts of the counties of Inverness, Argyll and the Lothians.

If it is accepted that the census in Scotland should reflect Scottish needs and circumstances in this way, we have to consider what subjects it should cover. The dilemma here is that, on the one hand, census users want information on as wide a range of topics as possible while, on the other hand, it would clearly be unreasonable to require every householder in Scotland to provide the whole range of information which in England and Wales will be spread across the 10 sample forms in such a way that each householder will be asked about some of the topics only.

Our starting point, as in England and Wales, is that in this mid-term census we ought to reduce the form-filling burden on the public as compared with that in the 1971 census. The Scottish Departments which use the census data in their work firmly prefer full enumeration of a somewhat reduced range of topics to a sample enumeration of a more extensive range of topics.

In order to reduce the Scottish questionnaire to a reasonable length, we have therefore had to make a judgment as to the topics on which we most need information. We propose to leave out some topics asked in England and Wales, not because the information would not be useful but because, in our estimation, it has a lower priority. So the overall range of topics covered in the single Scottish form will be smaller than the range spread over the 10 sample forms in England and Wales, but the Scottish form itself will cover rather more topics than any one of the 10 forms. In terms of topics, therefore, it will be slightly longer than any of the 10 English forms. But some of the topics will be subdivided into two or more actual questions and, reckoned on the basis of detailed questions, the Scottish form will be much about the same length as the longest of those south of the border.

As a result, the Scottish census will contain no new topics as compared with 1971, though some of the questions will be more detailed on this occasion, and there will be minor differences. Because of the differences in lay-out and the make-up of questions, it is difficult to compare the forms exactly for length, but the House may wish to know that, in terms of topics and questions, the Scottish form in 1976 is likely to be shorter than its counterpart in 1971 by nearly one quarter.

I hope that hon. Members will agree that we are right to shape the census to our particular needs, and that we have adopted a reasonable course in seeking to keep the demands on the public broadly comparable with those made in England and Wales, while reducing them from the 1971 level.

It may be helpful if I say a word about the reasons for excluding certain of the topics included in England and Wales. Those on which information will not be sought in Scotland are: the date when the person first lived at his current usual address; the countries of birth of each person's parents; the availability to the household of cooker, sink and hot water; school qualifications; and the place of study and journey to it.

I have already explained that, given the desirability of full enumeration and the need to keep the form to the essential minimum, we have had to judge what the topics are on which we most need data and what can be left out as having lower priority, however desirable they are in themselves. There is always room for argument about this, of course, but we have framed the list for Scotland with the needs of census users very much in mind.

Although we are not asking for the date when a person first lived at his usual address, we shall obtain on a 100 per cent. basis, information about people's usual addresses both one year and five years before census day, and this will provide us with two time-horizons against which to measure migration into each area in Scotland.

My hon. Friend explained the reasons for including in England and Wales the questions about the countries of birth of each person's parents. In Scotland we shall, of course, continue to ask about the country of birth of each person in respect of whom a census return is completed, whether adult or child, and so we shall be able to measure, in terms of the countries in which they were born, the composition of the population and the numbers who have moved between the different countries within the United Kingdom. But because the problems of the immigrant communities are less acute in Scotland—in 1971 we had only about 15,000 people both of whose parents were born in the new Commonwealth —we think that we can reasonably dispense with asking separately about the countries in which people's parents were horn.

I turn briefly to the important question of the availability of cooker, sink and hot water. We think that, in order to shorten the form, questions on this can be dispensed with in Scotland without real detriment to our knowledge of housing conditions. It is important to bear in mind that, in measuring housing standards, having the exclusive use of a bath or shower is more significant as an indicator than having the use of a cooker, a sink or hot water. The 1971 census showed that, of those households in Scotland which lacked exclusive use of at least one amenity, which numbered 231,000, in almost all cases—222,000—the lack was in the exclusive use of a bath or shower. We shall continue to ask about the availability of a bath or shower in 1976.

Are the Government doing anything about installing baths and showers? In other words, are they using the information from the census?

My hon. Friend can rest assured that account has been taken of the information that flowed from the census in respect of needs for baths and showers in houses, certainly in Scotland. I can speak only on behalf of the Scottish Office. Much of our housing improvement programme has been geared to the information we gained from the 1971 census. Those hon. Members who think that census information becomes available only after three years are mistaken in their belief. Census information begins to become available almost within three months after the forms have been completed, so information is continually coming forward almost from within three months.

I turn briefly to the school qualifications. We have a question on school qualifications which was first asked in Scotland in the census in 1971, and therefore that information is available. For this mid-term census we think it more important to obtain full information about qualifications at degree or professional or vocational level and this will be done on a 100 per cent. basis in Scotland.

I deal finally with the place of study and journey to it. This topic features on only one-tenth of the forms to be used in England and it requires three questions. The subject is generally of less importance in Scotland and, given the need to keep down the length of the Scottish form, I do not think it has sufficient priority to justify us in adding three questions to our single form in order to cover it.

In 1971 the Scottish public responded sensibly to the census, despite attempts to stir up trouble, and this must be recognised. We have little or no trouble and no lack of response in Scotland. We have every reason to believe that in the 1976 census the Scottish public will respond with the same degree of co-operation and sensibility that was displayed in 1971.

12.8 a.m.

No one can begrudge the time spent by either of the Ministers in explaining the proposals for the census or would wish to begrudge the time necessary for a thorough reply, if that is possible, to the points made in the debate. Nevertheless, it is perfectly clear from the course of the debate that the Government cannot, within the scope of one and a half hours, satisfy the very deep anxieties and serious questions which have been raised. Whatever may be your discretion, Mr. Deputy Speaker, as to the use of your powers under the Standing Order, I suggest to the Government that this is very specially a case where consideration by a Committee of the House would be valuable to all concerned. Whether this could be a topic which the Expenditure Committee could handle in the near future, or whether it could done by means of an ad hoc committee, I do not wish to define. But quite clearly, many of the matters which have been raised are more suitable to be dealt with by the hearing of evidence and cross-questioning than in the course of even an extended debate.

I want to raise only three points, two of them relatively minor. I was surprised that the Minister of State did not refer at all to the fact that there had been a report to the House on this order from a Joint Committee of both Houses —the Joint Committee on Statutory Instruments. That Committee drew attention to two points of definition which it believed might cause difficulty—the definition of a "British vessel" and the position of persons on oil rigs on census night.

No doubt there are answers to these points, but the answers ought to be on the record, and it is a little surprising that the Minister did not volunteer the answers.

My second query is about the British Isles or, rather, the United Kingdom as a whole. I think the Minister, perhaps inadvertently, at some stage used the expression "the British Isles". At any rate, some hon. Member did. The parent Act—the 1920 Act—was an Act for the taking of a census in Great Britain. It would be helpful if the Minister could indicate the way in which, and the powers under which, the corresponding census is to be conducted in Northern Ireland, which, together with this, will give the requisite picture for the United Kingdom as a whole.

My third point, of which I have given notice to the hon. Gentleman, is that I wish once again to draw attention to the very serious evidence of gross inaccuracy and inadequacy in the enumeration of immigrants of the first or second generation, particularly in areas where there are large concentrations. I can briefly indicate the nature of the evidence of this under-enumeration.

I should like first to take the case of the London borough of Newham. I have given notice that I intended to do so to the hon. Member for Newham, North-West (Mr. Lewis) who has personally taken a great interest in this problem. There are two series of figures which come into question. The first is what is called the schools count—that is to say, the count as at 1st January, year by year, up to 1973, of children in schools who themselves were born in the new Commonwealth or whose parents were born in the new Commonwealth, but omitting those who entered this country more than 10 years earlier. That is one set of figures. The other is the census which in the "small area 100 per cent. returns", gives the number of persons of the same description by age blocks. It is, therefore, possible to make an almost precise confrontation of the one figure with the other. The result will perhaps be clearer if I give the figures for Newham.

The census purported to show that in April 1971 in Newham there were 6,271 children of compulsory school age of new Commonwealth parentage on both sides.

That is the census figure. The schools count for January 1971—only three months earlier—gave 6,545. There will, of course, be a minor difference between the two figures because the school population is larger than that of compulsory school age, and there are one or two other minor reasons for variation; but we note that the school count shows, with due regard to these factors, a somewhat larger figure than the census total for children of compulsory school age.

However, the schools count omits all who themselves or whose parents were in this country 10 years before. Now, it so happens that the census gives us a division of the immigrant population—in the literal sense, namely, those who have immigrated into this country—into those who were here before 1961 and those who came after 1960, so that we know the proportion of potential new Commonwealth parents who were in this country before 1961 and whose children, therefore, who will have been excluded from the schools count. Those figures show in the case of Newham that one-quarter of all the new Commonwealth parents were already in this country before 1961. So we have an unaccountable absence of those parents' children, most of whom must have been at school on census day. The census figure is far too low to accommodate the full number of children attributable to the entire immigrant population.

This is not an isolated phenomenon. It is something which appears wherever there is a considerable immigrant concentration. I shall trouble the House further only by referring to four more London boroughs—Hackney, Haringey, Lambeth and Lewisham. In each of these boroughs the proportion of children of compulsory school age or at school is similar in the census count on the one hand and, on the other in the schools count, which omits—I fear I have said this several times. but it is the key point—all children who entered the country more than 10 years before, or whose parents did so. Yet we know from the census that in every one of those boroughs nearly half the immigrant population was already in this country before 1961.

That being so, it is simply incredible that there could be so close a similarity between the purportedly complete census figure and the deliberately incomplete school figure—incomplete by omitting the children attributable to nearly half the entire immigrant population. I wish to emphasise the point made earlier by the hon. Member for Harlow (Mr. Newens), that if we are to have these figures at all —and it is desirable that we should, otherwise we shall be prey to rumour and conjecture—it is important that their substantial accuracy should be placed beyond reasonable doubt.

I assert—I have brought this to the attention of the census authorities repeatedly—that there is here prima facie evidence that where there is heavy concentration there is severe under-enumeration. It would not be difficult to understand that that might happen. I wish to ask the Minister not to respond in detail to this point but to give an undertaking that the argument I have again put forward tonight, which can be substantiated in detail in case after case, will be seriously considered, with a check-back where possible, so that the matter will be settled once and for all, one way or the other, whether there is in this respect a major inaccuracy in the census.

12.18 a.m.

As a qualified statistician I want to refer to what we have heard about censuses versus samples, because there has been a lot of nonsense talked, in the other place and in this Chamber, about the subject.

A month ago I felt that there would be no objection to this order. I feel poacher turned gamekeeper tonight, because in the past month, having been interviewed by a local government official in a London borough and asked details not only about who else lived in my house but their names and various other personal and impertinent details, I begin to wonder what we are coming to.

We now have a census about to be put upon us and I have no confidence that the enumeration will be as confidential as it should be. I have found a Press statement from the Minister for Housing and Construction in which he says that he will be using the returns from the census to identify empty houses and under-occupied houses. These things are obviously something we are extremely concerned about in this House.

I am also concerned that the Government do not seem to have enough faith in their OPCS statisticians to believe that they can so rigorously organise their statistical sampling as to produce from a much reduced effort, at a much reduced cost, a thoroughly viable set of results on which future planning can be based. If the largest companies in the country and the largest organisations in the world can do this with accurate statistical sampling—if drugs can be put on the market after organised statistical sampling —I cannot see why we should go to the expense, in this interim period, of having a full-blown census which pries into people's lives nationally in the way that is already happening at a local government level.

The enumerators have freely admitted to those of us in the market research profession that they do not have sufficient training. They do not even have the selection processes to which I would subject an ordinary market research interviewer. I am completely dissatisfied with the White Paper and the evidence which has been presented to us in support of the need to have this full census. I hope that we shall get an adequate reply, but if we do not I shall have no hesitation in voting against the order.

12.22 a.m.

I am happy to associate myself with the right hon. Member for Down, South (Mr. Powell) in his criticism of the amount of time the Government have made available for the debate. They must be aware of the strong feelings on both sides of the House about this matter, and if the debate had lasted any longer there would have been further speeches criticising the Government's approach.

Criticism of the census and the whole principle behind it are not new. In 1753 an hon. Member introduced into the House an attempt to hold a census which was described at the time as
"subversive to the last remains of English liberty"
and as likely to result in
"some public misfortune or epidemical distemper."
I understand that although that attempt was successful in this House it was enthusiastically thrown out when it reached the other place and it was some 50 years before the first census was held.

The basic concern of the House tonight, and of people outside, concerns the confidence which can be expressed in the confidentiality of the census and the purposes to which it will be applied. A survey published in the previous Government White Paper after the 1971 census showed that about 35 per cent. of the respondents thought that the main purpose of the census was
"to provide a record of people's addresses for the government."
Fully 27 per cent. of those questioned thought that the information from individual census forms was available to
"people working for the government."
Clearly, when people believe such things, however wrongly, it illustrates the genuine concern felt throughout the country about invasion of privacy and associated matters.

The Under-Secretary of State for Scotland outlined the differences between the census for Scotland and that for England and Wales. The basic logic seems to suggest that there must be a different series of questions for the two censuses. Perhaps there is a glimmer of explanation, if not in the Under-Secretary's speech then in the White Paper outlining the Government's policy. In Paragraph 11, which deals with the number of rooms in accommodation, it states that the question on sharing is concerned to find out
"how many households have to share circulation space (or, in Scotland, rooms)".
It appears that what in Scotland is described as rooms is, in the rest of the country, mere circulation space. It would be interesting to know the basis of the distinction.

The basic criticism of the Government's approach concerns the question whether they have justified the need for a full census on a mid-term basis. This is the first time there has been a full mid-term census. Although the Minister made no attempt to explain the need for it in his opening remarks, the White Paper suggests the explanation why, with all the sophisticated mechanism provided by modern computer apparatus, we cannot use existing census information, whether on a 10 per cent. sample or otherwise to take account of the new local government boundaries. I do not regard the explana- tion we have been offered tonight as being good enough.

We know that the expense involved is considerable. In 1961 the census expenditure was about £5 million, in real terms. It went up to £10 million in 1971. The Minister told us that the £21 million which is estimated to be the cost of the intended census is not significantly different, in real terms, from the amount actually spent in 1971. If that is so, it suggests that the 1971 estimate was wide of the mark, and we are entitled to considerable doubt whether the estimate of £21 million is reliable.

While many of us accept that there may be good arguments for a mid-term census, we are not convinced of the argument that it should be a full census. We ask for an assurance that the fact that we are to have for the first time ever a full mid-term census is not to be taken as a precedent, and that we shall not be moving towards a situation in which, instead of having a 10-year full census, from now on we can expect a full census every five years. If the Minister believes that a full census is required this time because of local government reform, there may be a good argument for considering whether a further census in 1981 will be necessary. I ask the Minister for an assurance on those two matters.

I find it difficult to recommend my hon. Friends to support the Government, but I hope that an assurance will be given by the Minister both on this census and on any precedent that might arise from this debate.

11.26 p.m.

In the three and a half minutes available to me I shall not be able to reply fully to all the points that have been raised.

I shall comment first on what was said by the right hon. Member for Down, South (Mr. Powell). Having listened to the debate, I am in no doubt that the procedure which has been followed by successive administrations is not ideal. I shall consider seriously, in conjunction with my hon. and right hon. Friends, the possibility of improving the procedure. If there is any way in which it can be improved in the short term, I will see that it is done.

The expression "British vessel" is not defined. It is an expression which has to be given its ordinary or dictionary meaning. We realise that the expression otherwise could be ambiguous, although it has been used in two previous censuses. Oil rigs are not within the scope of the order, because the Continental Shelf Act does not make designated areas in which oil rigs are set up part of Great Britain.

No arrangements have been made for a census in Northern Ireland at present.

I have now dealt with the three points of detail raised in the debate. Under-enumeration is a complicated subject, and I will write in detail to the right hon. Member for Down, South about it. A great deal of work has been done which shows that the under-enumeration has been very small, but I shall deal with that in writing.

From listening to the Opposition Front Bench one would think that the Conservatives had never conducted a census, that they never realised that they were responsible for the 1971 census or that only a short time ago they rejected the idea of even having a White Paper on the subject.

The previous sample census was full of errors. We think that our sampling techniques have improved substantially. They are used in this census, but on the basis of a framework of an overall census. That is one reason why we think

Division No. 161.]

AYES

[12.32 p.m.

Beith, A. J.Mayhew, PatrickThorpe, Rt Hon Jeremy (N Devon)
Brittan, LeonMorrison, Hon Peter (Chester)Winterton, Nicholas
Chalker, Mrs LyndaPenhaligon, David
Durant, TonyRoss, Stephen (Isle of Wight)

TELLERS FOR THE AYES:

Hooson, EmlynTaylor, R. (Croydon NW)Mr. Cyril Smith and Mr. David Steel
Howells, Geraint (Cardigan)

NOES

Armstrong, ErnestFlannery, MartinMorris, Charles R. (Openshaw)
Bagier, Gordon A. T.Golding, JohnNewens, Stanley
Bishop, E. S.Hamilton, James (Bothwell)Oakes, Gordon
Canavan, DennisHarper, JosephOwen, Dr David
Carmichael, NeilHarrison, Walter (Wakefield)Prescott, John
Cocks, Michael (Bristol S)Jackson, Miss Margaret (Lincoln)Roderick, Caerwyn
Cohen, StanleyJohn, BrynmorRooker, J. W.
Concannon, J. D.Lamond, JamesSillars, James
Cox, Thomas (Tooting)Leadbitter, TedSkinner, Dennis
Craigen, J. M. (Maryhill)Lewis, Ron (Carlisle)Small, William
Crawshaw, RichardLoyden, EddieSnape, Peter
Cryer, BobLuard, EvanStewart, Rt Hon M (Fulham)
Cunningham, Dr J. (Whiteh)Lyon, Alexander (York)Stoddart, David
Davidson, ArthurLyons, Edward (Bradford W)Taylor, Mrs Ann (Bolton W)
Dormand, J. D.McElhone, FrankThomas, Ron (Bristol NW)
Dunn, James A.McNamara, KevinTierney, Sydney
Eadie, AlexMadden, MaxTinn, James
Edge, GeoffMarshall, Dr Edmund (Goole)Urwin, T. W.
Ellis, John (Brigg & Scun)Mellish, Rt Hon RobertWainwright, Edwin (Dearne V)
Ewing, Harry (Stirling)Molloy, WilliamWard, Michael

that sampling techniques are better. There is no question of the Government's rejecting advice. We are open to advice and rely heavily on the advice of our statisticians.

Many hon. Members mentioned cost. With the present stringency in public expenditure it is difficult to make a judgment on cost. Local authorities have supported the census throughout. If we were to run into considerable expenditure difficulties, obviously the Government would have to look at this again in terms of cost.

There is need for accurate information. For instance, population trends change in London. One needs information on forward population trends in planning district general hospitals, involving, in London, a capital of up to £20 million to £30 million.

It is used in terms of planning policy for housing—

It being one and a half hours after the commencement of proceedings on the motion, Mr. DEPUTY SPEAKER put the Question pursuant to Standing Order No. 3 ( Exempted business).

Question put, Thai the amendment he made:—

The House divided: Ayes 13, Noes 66.

Whitlock, WilliamWise, Mrs Audrey

TELLERS FOR THE NOES:

White, Frank R. (Bury)Woodall, AlecMiss Betty Boothroyd and
Whitehead, PhillipMr. Donald Coleman.

Question accordingly negatived.

Main Question put:—

The House divided: Ayes 66, Noes 13.

Division No. 162.]

AYES

12.40 p.m.

Armstrong, ErnestHamilton, James (Bothwell)Roderick, Caerwyn
Bagier, Gordon A. T.Harper, JosephRooker, J. W.
Bishop, E. S.Harrison, Walter (Wakefield)Sillars, James
Buchanan, RichardJackson, Miss Margaret (Lincoln)Skinner, Dennis
Canavan, DennisJohn, BrynmorSmall, William
Carmichael, NeilLamond, JamesSnape, Peter
Cocks, Michael (Bristol S)Leadbitter, TedStewart, Rt Hon M (Fulham)
Cohen, StanleyLewis, Ron (Carlisle)Stoddart, David
Coleman, DonaldLoyden, EddieTaylor, Mrs Ann (Bolton W)
Concannon, J. D.Luard, EvanThomas, Ron (Bristol NW)
Cox, Thomas (Tooting)Lyon, Alexander (York)Tierney, Sydney
Craigen, J. M. (Maryhill)Lyons, Edward (Bradford W)Tinn, James
Crawshaw, RichardMcElhone, FrankUrwin, T. W.
Cryer, BobMcNamara, KevinWainwright, Edwin (Dearne V)
Cunningham, Dr J. (Whiteh)Madden, MaxWard, Michael
Davidson, ArthurMarshall, Dr Edmund (Goole)White, Frank R. (Bury)
Dunn, James A.Mellish, Rt Hon RobertWhitehead, Phillip
Eadie, AlexMolloy, WilliamWhitlock, William
Edge, GeoffMorris, Charles R. (Openshaw)Wise, Mrs Audrey
Ellis, John (Brigg & Scun)Newens, StanleyWoodall, Alec
Ewing, Harry (Stirling)Oakes, Gordon

TELLERS FOR THE AYES:

Flannery, MartinOwen, Dr DavidMiss Betty Boothroyd and
Golding, JohnPrescott, JohnMr. J. D. Dormand.

NOES

Beith, A. J.Mayhew, PatrickThorpe, Rt Hon Jeremy (N Devon)
Brittan, LeonMorrison, Hon Peter (Chester)Winterton, Nicholas
Chalker, Mrs LyndaPenhaligon, David
Durant, TonyRoss, Stephen (Isle of Wight)

TELLERS FOR THE NOES:

Hooson, EmlynTaylor, R. (Croydon NW)Mr. Cyril Smith and Mr. David Steel.
Howells, Geraint (Cardigan)

Question accordingly agreed to

Resolved,

That items 7. 8, 9(d)(e)(f)(g)(h)(i)(j), 10,:1(a) (b)(c)(d), 14, 15, 17(a)(b), 18(vi), 19, 20, 21(a) (b)(c) in Schedule 2 and items 6, 8, 10(c)(d)(e)(f) (g)(h)(i), 11(vi), 12 13, 14(a)(b)(c)(d)(c), 16 and 17 in Schedule 4 to the Order in Council entitled the Census Order 1975, a draft of which was laid before this House on 20th February, be approved.