That the Grand Committee do report to the House that it has considered the Census (England and Wales) Order 2009.

Relevant Documents: 23rd Report, Session 2008-09, from the Joint Committee on Statutory Instruments and 29th Report, Session 2008-9, from the Merits Committee.

My Lords, this draft Order in Council gives effect to the proposals of the United Kingdom Statistics Authority for the ONS—the Office for National Statistics—to conduct a census throughout England and Wales in 2011. These proposals were set out in the Government’s White Paper, Helping to Shape Tomorrow, published in December 2008.

The census is the most important source of population statistics, and is the only source of statistics for both small areas and minority population groups. For over 200 years it has provided the underlying information which successive Governments have used to devise policies, take decisions and deliver services.

Billions of pounds of public funding and resources are allocated to local and health authorities each year using census-based information.

Since the 2001 census, a new independent body that reports directly to Parliament, the UK Statistics Authority, has been created. The ONS, the executive office of the authority, will lead on the census in 2011. I can therefore present the census order today confident that its content has been produced not by government but by the ONS, following extensive consultation with a wide range of stakeholders and with rigorous question testing. The independence of the census is essential if we are to build public trust in official statistics.

I will now take you through the detail briefly. Under the Census Act 1920, an Order in Council is necessary to prescribe the date of the census, the people to be counted, the people responsible for making a census return, and the information to be given in the census returns.

The date of the census will be 27 March 2011. Everyone will be recorded at the place where they are usually resident. Household members who are temporarily absent on census night also need to be counted. Some additional information will be collected on household visitors, which is essential if the ONS is to ensure that visitors are properly counted at their usual residence. The person responsible for making a census return is to be the householder or joint householder, but any individual aged 16 or over may make a separate individual return. The information to be given in the census returns is covered by Article 6 of the order and is set out in Schedules 2 and 3. Regulations, to be laid before the House in spring 2010, will set out the detailed content of the census and will contain copies of the various questionnaires.

On the information to be given in the returns, the ONS has carried out extensive consultations with central and local government, including members of all-party parliamentary groups, with community groups, with businesses, with academia, with the third sector and, most importantly, with the general public.

The ONS has also carried out in-depth question testing between 2003 and 2009, building on the valuable experiences and the lessons learnt from previous censuses. This is to ensure that the proposed questions in the 2011 census are justified, both in terms of the need for the information itself and of their acceptability to the public.

Under the terms of the Census Act 1920, some items—those in italics in the draft order—may only be included in the census if they are approved by an affirmative resolution. It is these that we are debating today. Only a few are new to the census, and I will now deal with these specifically.

First, people born outside the UK will be asked about their date of entry into the country and, for those who arrived in the past year, the length of their intended stay. There is an increasing need for more accurate and reliable statistics on migration in general, particularly on short-term migration. The census will therefore provide information on shorter-term migrants and temporary foreign workers, and help to give a better understanding of their needs and their impact on local labour markets and demand for local services.

Secondly, people will be asked to state the country of any passport that they hold. Government departments and the European Union require statistical data on the citizenship of the population. The question asks about passports rather than about citizenship. Asking directly about a person’s citizenship confuses respondents, who may mistake the question for one of national identity. People give much more consistent answers to a question about passports, with the result being easier to use to produce the necessary citizenship-related information.

Thirdly, people will be asked to register another address where they actually spend time. Such information will be particularly useful to local authorities which need to know the number of people who stay within their area and use local services during the week but who have a usual residence elsewhere.

Fourthly, people will record their national identity in addition to their ethnicity. Noble Lords will be aware of the amendment that was voted down in the other place to add a tick-box for “Cornish” in the national identity question.

Fifthly, people will be asked about their main language, allowing central and local government better to target language support and resources to those sections of society unable properly to access public services due to language barriers.

People whose main language is not English will be asked to state their proficiency in speaking English, with an additional, similar question about proficiency in Welsh in Wales. This will give a greater insight into the need for language training and community work to combat any possible discrimination and social disadvantage suffered by people whose language skills could be improved.

All other items requiring affirmative resolution cover topics that were included in the 2001 census, but I shall refer specifically to two where there have been some slight changes to the information to be collected. The question asking people to assess their general health during the past 12 months has now been expanded to a five-point scale ranging from “very good” to “very bad”. This information is for the planning of health policy and the provision of services, particularly for the elderly. The question on qualifications has a new tick-box to indicate any foreign qualifications held.

Other questions deserve mention either because the proposed wording has been revised since the last census or because particular interest has been shown by noble Lords. Following the Civil Partnership Act 2004, the traditional census question on marital status has been expanded to include response categories for civil partnership status.

Noble Lords may be aware of lobbying campaigns by Kashmiri, Sikh and Cornish groups, for example, on the ethnicity question. It has not been possible to include a tick-box for every ethnic group category requested. Each request was compared against a detailed set of prioritisation measures before the final recommendation was made. This identified strongest need for additional response categories for “Gypsy and Irish Traveller” and “Arab”, which have therefore been included for the 2011 census.

People can record their ethnicity in whichever way they choose by writing this in the write-in spaces provided. All such responses will be counted. In the case of Sikhs, there is the additional option of ticking the specific “Sikh” box in the religion question. The Government are satisfied that the census is compliant with national legislation on race relations and any international conventions on the protection of national minorities.

I shall refer briefly and specifically to the proposed question on religion. A number of noble Lords have made representations to the House during previous debates and directly to the ONS on the proposed question. In particular, concerns have been expressed that the question on religion is leading and overstates the extent of religion within the country. The proposed question measures an affiliation with a religion and not belief or practice. It provides information which can be used together with responses to the ethnicity question to monitor equality and discrimination by identifying minority ethnic subgroups that may otherwise be missed, particularly those originating from the Indian subcontinent, in terms of their religion.

As a result of concerns expressed, the ONS tested alternative questions on religion. The testing concluded that no single question on religion can be worded in such a way as to capture suitably information on religious belief or practice and to justify losing comparability with the 2001 census. There is a strong user requirement for comparability with the data from the 2001 census. The proposed question on religion therefore remains unchanged, and remains the only question in the census for which a response is voluntary.

There is a new question asking for the number of rooms designed for use as bedrooms. This will measure overcrowding and assist in the planning of housing investment and the addressing of social and economic need. There is also a new question asking about the type of fuel used for central heating, to provide data on basic housing standards and fuel poverty and deprivation.

Despite some strong user demand for an income question, testing by ONS revealed that including a question on one’s income leads to a significant drop in response rates. There were also concerns over the quality and accuracy of the income data collected as part of the test.

There was a strong requirement for information on sexual identity in order to monitor equality. After investigation, the ONS recommends that a compulsory census is not a suitable means to collect information on sexual identity, again because of the quality of the resulting data, and including the question is likely to reduce the overall response to the census. Although the ONS does not recommend that a question on income or sexual identity be asked on the census, it recognises the user demand for this information and is taking steps to provide some data through other sources.

I shall say a few words about security and confidentiality. The ONS has an excellent track record when it comes to security. Additional measures are being instigated and all internal and external threats that could affect the confidentiality, integrity and availability of census information are being carefully managed. The information will be protected by strong physical and electronic data security arrangements.

Census information will be used only for statistical purposes, and ONS will not release any personal-level information. Noble Lords will know that it is a criminal offence to unlawfully pass on personal census data. Census information is exempt from disclosure under the Freedom of Information Act and the Data Protection Act.

The Government’s policy is that data from questionnaires should be kept secure and closed to the public for the next 100 years, with the paper questionnaires being destroyed in a controlled and secure manner. The resulting electronic data will be registered under the Data Protection Act 1998, and the arrangements for handling the information during processing are to be the subject of an independent security review.

I turn to the cost of the census. The 2011 census in England and Wales is expected to cost around £480 million over the 10-year period. In its report on the 2001 census, the Treasury Select Committee recommended that any future census should be justified in cost-benefit terms. The ONS business plan to the Treasury for the 2011 census clearly demonstrated the unique value of the census and that the benefits of having the information far outweigh the costs of its collection. The identified cost benefits exceed £700 million, compared to a cost of £480 million.

More than £1 trillion is allocated to local authorities and to NHS primary care trusts in the 10 years between each census, on the basis of census information. The census cost equates to just 87p a year per person, which we believe demonstrates excellent value for money. The per capita costs in the UK are less than those for many other European countries that carry out similar censuses, and considerably less than the costs in America.

In summary, this census will meet crucial requirements for the statistical information that Governments and others cannot do without. The burden on respondents will be kept to a minimum and the questions have been fully tested and found acceptable and compatible with both UK and EU legislation. I am confident that we have the right census package for helping to shape tomorrow. I therefore commend the draft order to the Grand Committee. I beg to move.

I thank the Minister for that thorough presentation of the detail of the order before us. I can keep my remarks brief in response as many of the issues were covered this week, as I am sure the Minister is aware, in the debate on this issue that took place in the First Delegated Legislation Committee in another place. That debate raised a number of the issues which we would otherwise have wanted to explore today, but I do not wish to detain the Grand Committee too much.

A number of questions were asked by my honourable friend Nick Hurd in the First Delegated Legislation Committee, but I wish to take this last opportunity to scrutinise the order and to elicit responses from the Minister, if possible, to his fair and reasonable questions. I should like to go over again the questions on which we seek clarification.

The first question relates to costs, to which the Minister has just referred. The costs of the census in 2001 were £214 million; it is projected that the costs will rise to £482 million in 2011. Clearly this increase in costs represents a significant increase over the rate of inflation. As not many more homes will be covered, and given that the public finances are in a state of distress, would it not be appropriate to revisit that figure to see whether it can be reduced? However, before pursuing that line, can the Minister respond to the specific question of whether the increase in costs can be explained?

The second question asked by my honourable friend Nick Hurd in the Committee in another place concerned the assumptions used in awarding the contract to the American company Lockheed Martin. What assumptions were made about the response rate for the 2011 census? Some data show that the census response rate in 1991 was 96 per cent and that in 2001 it had fallen to 93.9 per cent. An essential critique of the order before us is that, by lengthening the form and increasing the number of questions and their potential complexity, the level of response might be lower, and presumably when the contract was awarded to Lockheed Martin an assumption was made. Could that assumption be put on record?

During the previous census there were seven times more calls than expected to the helpline which had been set up to assist people in completing the form—2.6 million in total. Given that the new census form is longer and more complex, what assumption has been made about the number of calls that might be made to the call centre, and what steps are to be taken to ensure that the extraordinary delays which people experienced previously are not experienced this time round?

This will be effectively the first census carried out during the broadband era, if one might loosely describe it as such, and a lot of people will use computers to submit their response. There was e-mail help and technical support in 2001 but that system, too, was inundated. Can the Minister say what the response is likely to be in 2011 and what provisions have been made to ensure that people receive the help that they need with the forms to ensure as high a completion rate as possible?

The next question relates to the overall costs of those two factors, and this was one of the assumptions to which I should be very interested to hear the answer. What assumption is being made about the number of respondents who will complete the form online? That is an interesting point because my intuitive judgment is that the figure will be significantly higher than was the case in 2001. If that is so, again, intuitively one would expect that if more people complete the form online, there will be a need for fewer enumerators to knock on doors and fewer pieces of paper to be printed and processed, and that should surely reduce and not lead to a doubling of the costs.

Fourthly, my honourable friend Nick Hurd asked whether the Minister could publish the response rates to the 2009 rehearsals which the ONS and the Government conducted in a number of areas. This would place important findings in the public domain and would, had they been available, have informed the debate and perhaps allayed some unnecessary fears. Some have suggested that if the Government had found the response rates and the responses during the 2009 rehearsals helpful to their case, they might have deployed information received from them in support of these orders. Therefore, the fact that no information has been made available to the House on those rehearsals raises some concerns about confidence.

Fifthly, can the Minister tell the Committee roughly what proportion of responses will be extrapolated from fields that have been left blank? I understand that statistically what happens is as follows: if all 36 questions on the form have not been responded to—if, for example, five have not been answered—the enumerator extrapolates or guesses what the response would have been. Clearly, given that the Government are placing a great deal of store on the accuracy of data, if a significant amount of the data is the result of guesswork, that will raise a serious concern. Will we be guessing 20, 30 or 40 per cent of the answers? If, for example, as we were told, more than £1 trillion of taxpayers’ money is to be directed by these data to health and local authority spend alone over the next 10 years, can we be confident that the data informing these choices are correct? Is it possible that the cost arising from misdirecting and misallocating funds could be much higher than the cost of the exercise itself?

Does the Minister agree with some of the expert opinions that have been expressed about the complexity of the data? Is she aware of the remarks of Professor Rees, a leading UK statistician, to the Public Accounts Committee on 19 November, when he said that as you expand the number of questions there must be a deterrent effect on response? Has she noted those comments? Could she respond to them?

My Lords, I thank the Minister for giving us a clear exposition of the importance of the census as an indicator of factual information that is utterly vital to determining, among other things, the distribution of public expenditure. I also suggest that this is quite an important debate, because, as the noble Lord, Lord Bates, said, it is the last public opportunity for consideration of these issues before the process goes into its concluding stages. He raised many points which I shall largely take as put so that the Minister can focus on them, but I have a number of questions and points of my own.

The consultation to which the Minister referred has been extensive, which is entirely appropriate, and I am glad that it has been conducted in the way that it has. It would be of interest to know whether the stakeholders to whom she referred, particularly the institutional stakeholders, had any comments to make about the costs. I was slightly surprised to read in the Explanatory Notes that no impact assessment had been produced for the order since no impact on the private or voluntary sector was foreseen. I am surprised not least because of the responsibility of those in communal establishments for providing information or for ensuring that information is provided. That includes hospitals, care homes and prisons—although they are not private institutions. There are cost implications for those complying with it. I accept that, for individuals, half an hour would probably be a sufficient allocation of time to reply even to the enlarged questionnaire.

The second broad policy question which I would like to raise—I am not aware that it has been raised in earlier debates—concerns the frequency with which the census is held in this country. There has been a 10-year census for a long time, I think, since it was introduced in 1801. However, I have noticed that a substantial number of countries, including Australia, Canada, Eire and New Zealand, conduct a census every five years. I do not know what extra cost that involves, or indeed whether some of the costs in the 10-year census might be obviated by having one more frequently, but it is fair to say that we are living in a very rapidly changing society. To some extent, that is reflected in the enlargement of the questionnaire to four pages from three—a 25 per cent increase—in the 10-year census. If we are making judgments about the statistics which it will yield, perhaps we should think about whether a census ought to be conducted within a shorter timeframe in future, so that the responses of Government to many of the issues to which the Minister referred could be more immediate and less “after the event”.

The cost, which has been extensively referred to, is not inconsiderable. The Minister referred to the cost benefit, but £482 million is a very substantial sum. I would be grateful if the Minister would say a little more about the extent to which competition has been, and can yet be, involved in conducting the necessary preparations for this census. The competitiveness of contractual offers should be borne in mind, as well as the robustness of the testing of the possible conduct of the individual parts of the exercise. The census has produced a highly valued series of statistics historically, and this country has a good record on that which is generally recognised. It would not be proper to imply that I am seeking in some way to cut corners. Nevertheless, with very substantial contracts of this kind, it would be helpful to know about the process—who conducted it and how it was scrutinised—not necessarily because changes will be made on this occasion, but perhaps if they have been made in the process which has run in the 10 years since the last census.

As far as the range of questions is concerned, I have no great problem with the enlargement of the number of questions in principle because, as I have said, there have been considerable changes in the mobility of our society, in the ownership of second homes, and in the movements into and out of this country. The only question about that latter point is whether the personal information being sought in respect of visitors is really required or whether it is unduly intrusive. I have no doubt that the Minister will have a view about that.

I am sorry I was not here to listen to the first few paragraphs of the opening remarks of the noble Baroness. I was very interested in the remarks I did hear and the speeches of both noble Lords opposite. I suppose I should be proud to be the only representative of the Back-Benchers in the House of Lords to attend this erudite debate.

I do not have too much to say. but I wish to raise the issue of confidentiality and the Freedom of Information Act. The noble Baroness assured us that the security could not be breached, which one has to accept. However, I remember that during the problems—if I can call them that—in the House of Commons over expenses, the Members of the House were assured that they were covered by privilege. However, it turned out that they were not covered by privilege; the information they were given about the security of the details, which they had given in confidence and expected would not be revealed to the general public, proved to be completely and utterly wrong and has caused certain problems. I hope the noble Baroness can reinforce the assurance she has given this afternoon that the information is definitely confidential and cannot be released to anyone.

My second point is on ethnicity.

Before the noble Lord moves on, is he interested in finding out a little more about Lockheed Martin, an issue raised by the noble Lord, Lord Bates, and, in particular, what constraints there are on the information to which Lockheed Martin will have access? What will restrain it from losing, copying or otherwise mishandling the census data, bearing in mind that some may consider it is bound by the Patriot Act in the United States?

That is a very good intervention and very helpful. We need such assurances because so much information now is put on disks or whatever other storage capacity is available, such as memory sticks. Bearing in mind that some very important information regarding people’s taxation has been lost, we are entitled to know what security is available and whether a private firm is as secure as I hope our own departments are. We need those important assurances.

I will move on to ethnicity. The noble Lord, Lord Maclennan, raised this, and it was right and proper that he did. We are becoming obsessed with two things: ethnicity and sex, which I will come to in a moment.

The drive in this country should be for unity; instead of multiculturalism, we want integration. We want people to be happy that they are in Britain. We want people to feel British, and to feel that they belong here, rather than anywhere else. They should feel that they are a part of us, no matter what their colour, creed, religion, or anything else. I feel strongly that the motivation of Government, and indeed of Parliament, should be towards integration. I am not at all sure that asking people to define their ethnicity does that.

My wife and I had a recent experience with our local authority in Reading—not about the census, but about the electoral registration form. We decided to test the market, so to speak, to see what would happen. I put my nationality down on the form as Welsh, and my wife said she wanted me to put hers down as English. Lo and behold, when the register was proclaimed and I got the form for next year, there it was. My nationality—Welsh—was fine. I could be Welsh, but my wife could not be English. That was very worrying. The local authority would not accept “English”, so it put “England”. This, of course, was a ridiculous answer to the question of nationality, because there is no such nationality as England. For three years, I have been trying to get the local authority in Reading to accept English as a nationality. I still have not succeeded—they keep blaming it on a computer. I do not know whether English is to be accepted as a nationality by all authorities now, bearing in mind that it is on the census as such, and I would be interested if the Minister could comment on that. Perhaps it is a little unfair to ask her to do so, but nevertheless if she could I would be most obliged. My wife is very offended, and I am sure there are many other people who are offended that they cannot be English. As a Welshman, I am standing up this afternoon for the English.

I am extremely grateful to the noble Lord, Lord Stoddart, for allowing me to intrude again. It is just conceivable that there are people, and I would certainly count myself one, who regard themselves as glad to be Scottish, British, and European. I do not expect the noble Lord to be enthusiastic about that proposition, but it might be of interest to have such a multiple identity disclosed in the census returns.

I have to tell the noble Lord, Lord Maclennan, that he is okay. The electoral registration officer will accept that he is Scottish. The trouble is that they will not accept that my wife is English, and she does not like it. There are a lot of other people around of the same opinion.

As for being European, I consider myself to be part of the continent of Europe—we are, after all, on the same continental shelf, which I am very happy about. However, I do not want be governed by an organisation called the European Union, which is completely undemocratic—I shall not go into its problems—and I am quite happy to be Welsh, British and European all at the same time. We should move towards an integrated community, not a split one.

I am glad that the question on income has been dropped. It is clear that there was a lot of discussion. I think that the Government perhaps favoured our being forced to answer questions on it, which would have been very difficult. A lot of income comes from the black economy. You cannot really expect people to put, “This is my income from work but then, on the other hand, I do some other jobs on the side and this is how much I earn”. I am not sure, therefore, that one would get an accurate view of people’s incomes. People have all sorts of income, some of which they forget or do not want to put on a tax return. They will not fill in a census which, although they are told it is confidential, may lead to the taxman getting hold of some income that they have not declared to him. I am glad also that the question on sexual proclivities has been dropped, because there are so many of them. It would require a very long list of questions to get a true answer.

However, I am concerned that these intrusive questions should not be put about by other means. While they will not be included in the census, it is clear that there is curiosity among the Government about people’s incomes and how they behave themselves sexually. I would like the Minister to give me the relevant assurances.

My final point echoes one raised by the noble Lord, Lord Maclennan. Our population since the end of World War II has risen by more than 15 million, much of it taking place in the past five years. Consideration should be given to taking censuses at shorter intervals. It might not be as dear as one would think, because a lot of surveys about various things are being done up and down the country in the interim, and they cost money, too. While we might on the one hand incur additional costs because the census would be conducted at shorter intervals, we could save money on the other because there would be less need for individual surveys on various items. In the hope of one or two assurances from the Minister, I shall sit down.

I thank noble Lords for a robust, vigorous and detailed discussion. That was as it should be because, as the noble Lord, Lord Maclennan, said, this is an extremely important subject. Like the noble Lord, Lord Stoddart, I am a little disappointed that more Members of the House were not interested in it, because, as the noble Lord, Lord Bates, said, the census will influence not millions or billions but trillions of pounds of spending by the public sector during the next 10 to 12 years. So it is big stuff.

I start by saying that I have good news for Lady Stoddart, which is that she will be able to put “English” as her nationality on the census. I hope the noble Lord will be able to give her that news. I ask noble Lords for their patience with my replies because there is quite a bit to get through.

The noble Lord, Lord Bates, asked about costs and whether there was a possibility of a reduction in them in the light of the economic climate that we are experiencing. Costs have been drawn very tightly by the ONS and its contractual programme. It has already spent 20 per cent of that sum of £480 million, 50 per cent is further committed to fixed-price contracts and any further reduction would impact on the quality of the census population estimates. I go back to the point, which we have all appreciated, that the quality of these statistics will influence an enormous amount of spending over the next decade. Therefore, the quality has got to be right. We must not skimp or cut corners, as the noble Lord, Lord Maclennan, put it, on quality—it is the main point that we have to keep our eye on.

My question was to ask why the cost had increased from £211 million to £482 million. What was the justification for that?

The short answer is that the main reasons for the increase are the extra £150 million due to inflation, the £25 million due to growth on numbers of people or households, the £25 million for the extra page of questions for each person, the extra £45 million for improvements in field systems and processors—those actually working on the 27th, making sure that the hard-to-reach people are visited and that censuses are delivered and sent back—and the extra £20 million for improved help facilities, publicity, community and local authority liaison, with a variety of other smaller changes and improvements. I hope that answers more fully the noble Lord’s question.

The noble Lord, Lord Bates, also asked about the justification for the cost as a general question. The full 2011 census business case has been produced, which will fully justify the census investment. For a subset of the different uses, the benefit has been assessed as exceeding £700 million. This is thought to be a significant underestimate of the total benefits.

The noble Lord, Lord Bates, talked about the assumptions that were made about the level of response when awarding the contract to Lockheed Martin. We believe that we will get a 94 per cent overall response, and that was the assumption when awarding that contract.

The noble Lords, Lord Bates, Lord Maclennan and Lord Stoddart, raised the issue of security and the awarding of the contract. The contract for the census was awarded to Lockheed Martin UK as it offered the best value for money in an open-procurement scheme carried out under European Union law and EU procurement directive. The company has a good track record on census work, having supported the 2001 UK census. All persons acting in any capacity for the UK Statistics Authority in the conduct of the census will be covered by the same stringent confidentiality restrictions as are applied to members of the Office for National Statistics themselves. There are severe penalties, as the noble Lord, Lord Bates, was saying, for breaches of census confidentiality laws. No Lockheed Martin employees will have access to the data.

Presumably a department of state—the Office for National Statistics, for example—was given the opportunity to tender for the contract.

I am happy to write to the noble Lord about this, but it is not our belief that the ONS could tender for the sort of work that Lockheed Martin will do. I shall go into that a little more. Lockheed Martin UK is the prime contractor. It has engaged a number of specialist UK and EU subcontractors for the different components of the contract. The main ones are Polestar for the printing—so the ONS would not necessarily be appropriate there—UK Data Capture for the scanning and the data processing, Cable and Wireless for the communications and the data centre, bss for the contact centre, Logica and Steria. I hope that answers the noble Lord’s question.

If the company is going to be responsible for the customer contact centre, how is it not going to be in touch with information that is going to be contained in the census?

The ONS has put in place additional contractual and operational arrangements in this contract to ensure that the US authorities could not gain access to census data through, for instance, the Patriot Act. These arrangements include that all data processing will be carried out in the UK, no data will leave or be held at any point outside the UK, all data are the property of the ONS and only UK/EU-owned companies will have access to the personal census data. The only people to have full access to the full census data set in the operational data centre will be the ONS staff.

The noble Lord, Lord Bates, asked about response rates for the recent census rehearsal. The responses are still coming in so the number is increasing, but at the moment they are at around 35 per cent. This is lower than hoped, although we have to remember that this, unlike the census, was a voluntary exercise. The ONS anticipated a much lower response than we will get in 2011, because a compulsory response will be much higher. After a promising start with response rates ahead of expectations, returns are now levelling off and lag behind expected levels. But the rehearsal is voluntary, and known hard-to-reach areas were selected for the rehearsal, which also tips the result. The exercise was not spread across, as it were; those areas were selected for the rehearsal as being hard to reach.

That was masterful. It was a wonderful answer to a series of complicated and focused questions. I am very grateful to the Minister.

I thank the noble Lord. The noble Lord, Lord Maclennan, asked why there will be no impact assessment. As no business or voluntary body will be required to undertake any duty that the ONS believes might prove detrimental to their work, delay their normal day-to-day business or place an additional cost upon them other than asking the manager of any commercial establishment to hand out or collect questionnaires to any usual residents and then complete a short questionnaire in respect of their establishment, the ONS believes that the 2011 census impact on the private or voluntary sector will be minimal. The noble Lord talked about it taking half an hour to complete a form; I have seen in my briefing reference to 10 minutes, but we will see. An impact assessment will be published before the census regulations are laid in the spring of 2010, so there will be one before the regulations.

Other countries have a five-yearly census, and the noble Lord highlighted those countries. Given our highly mobile population, particularly considering the money being allocated, why would we not have a five-yearly census? There were proposals for a five-yearly census in 1976, 1986 and 1996, but they were all rejected on cost grounds. The ONS is currently reviewing the option beyond 2011, so there will be a review of how we take it from here, given the issues that the noble Lord raised about the great changes in our society—the electronic changes, the ethnic changes and the way our country looks now compared with how it looked 10 years ago. There will be a review of whether a 10-yearly census is the way forward.

The noble Lord, Lord Maclennan, asked about stakeholders being consulted on costs. Stakeholders were widely consulted on the pros and cons of different options. Population-estimate quality is paramount. This cost and this level of investment are required to provide data of a profoundly excellent quality.

The noble Lord asked about competition for preparations in order to reduce costs. Fifty per cent of the census costs are for outsourced services, all of which have been subject to full competitive tendering. If the noble Lord wants more information on that, I shall be happy to correspond with him.