NINTH SPECIAL REPORT
The Treasury Committee has agreed to the following Special Report:
THE 2001 CENSUS IN ENGLAND AND WALES:
GOVERNMENT RESPONSE TO THE COMMITTEE'S FIRST REPORT
OF SESSION 2001-02
- On 6th March 2002, we published our First Report, The 2001 Census in England and Wales, as House of Commons Paper No. 310. The Report originated on our Sub-committee. We have now received a response from the Government (dated 22 May) which is appended below.
On the need for a Census
(a) The Office for National Statistics (ONS) were unable to supply us with robust evidence to justify expenditure of over £250 million on the Census. We recommend that a rigorous cost-benefit analysis should be carried out of the 2001 Census and published in time to inform the need for, and timing of, any future Census. We consider that any future Census should also be justified in cost-benefit terms. (Paragraph 10)
The prime justification for the Census is the well recognised need to understand the population, not only counts of people in the groups important to public policy and planning of services to communities, but the structure of the population itself.
A Census is a massive investment, with a very long time frame. At the time when it was necessary to begin planning the 2001 Census, it was judged that we had no options of significance that could replace the 2001 Census and meet the key needs.
The Government accepts the recommendation that the case for any future Census should include a cost-benefit analysis of the options available. This will draw upon the assessment of benefits and costs of the 2001 Census that has already taken place and been considered by Parliament, the Treasury and others, including:
1. The 1992 Census Policy Evaluation and Reappraisal undertaken by the three United Kingdom Census Offices. On the basis of this analysis the then Government decided that:
- there would be a continuing need both in the public and private sectors for the type of statistical information on population and housing provided by the Census over the period 1996-2016; and
- there were no alternative sources for providing such information on a nationwide standard basis for local areas.
2. The business case for the 2001 Census, approved by HM Treasury in 1998;
3. The analysis set out in the 2001 Census White Paper (Cm 4253) published in March 1999;
4. Parliamentary debate on the Census Order (by a House of Commons Standing Committee on 2 February 2000 and by the House of Lords on 16 February 2000).
The methodology for the review will be developed in discussion with Census information users, the Statistics Commission, international experts and ONS auditors.
(b) We also recommend that ONS publish estimated income to be received from all sources for the use of 2001 Census data. We recommend that their charging policy be published and regularly reviewed. (Paragraph 11)
For the 2001 Census access to all standard statistics will be through a user-friendly web site and, subject to simple licence terms, access to the web site will be free at the point of use. For non-standard statistics, Section 4(2) of the Census Act requires that customers will be charged only the marginal cost of producing and disseminating the information. This is also consistent with the Government's wider policy on the 'Knowledge Economy' to make best use of information held by Government by offering an extended range of data free of charge, thereby demonstrating its commitment to ensure that data is widely available and easy to access.
Full details of the range of Census outputs, means of access, charges and costs, terms and conditions of use, and planned timetable for release, were announced by the ONS in a Census Output Prospectus issued in March this year.
(c) We consider that in evaluating the benefits of any future Census all alternatives should be considered, including doing without a Census altogether and reverting to a simple headcount. Any evaluation will also need to take account of the likely acceptability to the general public of drawing on other data sources, such as tax and benefit records, for Census purposes. The Committee would wish to be informed of the results of these evaluations, which will need to be completed before a decision is taken on whether to hold a mid-term Census in 2006. (Paragraph 15)
The Government accepts the recommendation.
The next 'Census' is likely to be some mix of administrative records, household survey and perhaps a national count. The practical aspects of access to administrative records of high quality are the most likely determinant of the mode of the next Census.
The review outlined above will form the basis for justification of any future Census.
ONS has already carried out a separate review of the need to conduct a Census in England and Wales in 2006. The Government is currently considering the view of the National Statistician that there is insufficient justification to conduct a Census in 2006. An early decision on this will be necessary since any activity necessary to put into operation a Census in 2006 would need to start in 2002.
On the preparations for the Census
(d) We believe the legislative procedure for the Census provides a valuable opportunity for Parliament to make a more measured input to proposed legislation than when, as is usually the case, the enabling legislation permits only approval or rejection of proposed subordinate legislation. (Paragraph 19)
The Government welcomes the view of the Committee that the legislative procedure provides both a sufficient framework to conduct the Census and the opportunity for Parliament to express its view.
(e) It is clear that a question on income would have been found useful by many users of Census data. The ONS decision not to include such a question appears to be based, at least in part, on judgements on the public's willingness to provide such information in the context of a Census. The information on the response rate to the question on religion suggests that compulsion is not necessarily an essential element in securing a high level of responses to specific questions. We therefore recommend that, in the light of the response rate for the question on religion, ONS should consult further on whether a question on income should be included in any future Census, and whether it should be voluntary or compulsory. (Paragraph 23)
The judgement about whether to include a question on income was a difficult balance. The key criterion in assessing the success of a Census is the level of population coverage.
In a major test of Census questions and procedures in 1997, there was a statistically significant difference in the overall response rates between those households who had a form with an income question and those who did not.
ONS will consult further on whether a question on income should be included in any future Census.
On the issue of voluntary and compulsory questions, under the provisions of section 8(1) of the Census Act 1920 it is an offence for someone to fail to comply with the Act or any Order or regulations made under the Act by refusing to make a Census return. There is no legal provision for the inclusion of a question on a voluntary basis other than that for the specific question on religion.
(f) We note that ONS's preparations for the 2001 Census did not reveal the latent concerns in Wales over how Welsh identity was to be recorded on the Census form. We recommend, in the light of this, that ONS look further at the way in which they rehearse any future Census, to ensure the most accurate picture possible is obtained of both how the Census is likely to be received and proceed. We endorse the proposal that the National Assembly for Wales have a more formal role in agreeing future Census forms for Wales. (Paragraph 27)
The desire to be able to record 'Welsh' identity in a pre-defined response tick box followed the addition by the newly established Scottish Parliament of a 'Scottish' response category to the Scottish Census form, after the form for England and Wales had been agreed by Parliament.
Developing this interest in measuring national identity is most significant in devolved countries, but is also present to a lesser extent in England, and within England, for example amongst the more distinct communities such as the Cornish. Collecting information embracing this broader sense of identity is unlikely to be settled by simply changing the ethnic question on the Census.
Preparation for any future Census will include a more formal role for the National Assembly for Wales. The Government welcomes the Committee's endorsement of this approach.
(g) We welcome ONS's recognition that more needs to be done for blind and partially-sighted people for the next Census. We believe that this should mean that blind and partially-sighted people are able to take part in any future Census independently. More generally, we recommend that ONS review carefully, in consultation with relevant organisations, problems experienced by disabled people in relation to the conduct of the 2001 Census and adopt a best practice approach to meeting the needs of all disabled people. (Paragraph 31)
The Government accepts the recommendation and will review, as part of its evaluation programme, aspects of dealing with the visually impaired and other disabled groups in order to adopt a best practice approach to meeting the needs of all disabled people.
On the conduct of the Census
(h) In view of the difficulties experienced with the return of Census forms, we recommend that for any future Census ONS evaluate the benefits of postal return versus enumerator or other means of returning forms. If a future Census is to be based on postal return, ONS should conclude a service level agreement with the service provider aimed at ensuring that the conduct of the Census is not impeded by the quality of service received. (Paragraph 36)
As part of the review into the methods for the return of forms ONS will take fully into account the experience of the arrangements with the Royal Mail in 2001, and will ensure that future agreements with providers of any postal return service will address any such issues that may affect the quality and speed of the response.
(i) ONS seriously underestimated the volume of calls to the Census helpline and the service was initially overwhelmed. As a result, the service provided to the public, particularly in the week preceding Census Day, was inadequate and unsatisfactory. We note the significant number of calls for help received by ONS, but also recognise the context that, at some 2.6 million over three months, the number of calls received was under 10 per cent of those being enumerated. We recommend that helpline capacity for future Censuses be carefully evaluated, and that there are discussions with service providers to seek to minimise technical constraints on call volumes and handling peak loads as these factors clearly contributed to the difficulties faced by some callers. (Paragraph 40)
The helpline was initially overwhelmed, but the service provider and ONS were quickly (within 2 days) able to provide the required level of service. While the initial estimates were wrong, we regard the responsiveness as a considerable success, given that the use of the helpline reflected an overwhelming goodwill of the public to participate in the Census.
The Government accepts the recommendation and will review the performance of the helpline as part of its evaluation of publicity and communication issues.
(j) We consider that the number of calls received on the helpline suggests a need to make the Census process as simple as possible and for improvements in the way the public is informed of the Census and how to complete the form. We recommend that ONS review their communications campaign and the nature and volume of the calls received on the helpline to determine what can be done to reduce the level of help sought by the public in future. (Paragraph 41)
We consider that the scale of the public response to the publicity campaign demonstrates a good deal of success in heightening the awareness of the population to the Census that was about to take place.
The Government accepts the recommendation and will review the communication campaign and the nature and volume of calls received on the helpline to determine whether more can be done to make the Census process as simple as possible and reduce the level of help sought by the public in future.
(k) We note that the available evidence suggests that the quality of the Census data collected in the countryside has not suffered as a result of the foot and mouth epidemic. We congratulate ONS on reacting sensibly and pragmatically to an unexpected but very serious problem and the field staff on their sensitivity to the special requirements of the farming community. (Paragraph 44)
The Government welcomes the comments of the Committee. It acknowledges that the success of the Census in rural areas in the face of the difficulties imposed by the foot and mouth outbreak, was due to a large extent to the diligence and perseverance of the locally recruited temporary field staff, most of whom would have been familiar with, and sympathetic to, the local conditions prevailing at the time. The co-operation of the public in responding so well in such areas and the efforts of the Census service providers involved, most notably the Royal Mail and TNT, should also be acknowledged.
(l) We note the very low number of cases that have been referred for prosecution for failing to complete and return a Census form and we question what message this sends regarding the seriousness with which this offence is regarded. We recommend that ONS review the reasons for the very low prosecution rate. We also recommend that, in undertaking such a review, ONS research public attitudes to participation in the context of a mixture of compulsory and voluntary questions, as was used in the 2001 Census. (Paragraph 46)
ONS will conduct a full evaluation of its non-compliance policy and procedures once the programme of prosecutions is completed.
ONS has to date followed a policy of only pursuing cases where it has clearly sufficient evidence of a persistent refusal that more or less guarantees success in the Courts.
As noted above there is no legal provision for the inclusion of a question on a voluntary basis other than that for the specific question on religion.
On the results of the Census
(m) We recommend that in their evaluation of the 2001 Census, ONS review the balance of resources devoted to enumeration in the best performing areas and those devoted to the worst and consider what changes may be necessary to the Census in the light of the response rates to individual questions. In view of the comments of the Market Research Society, ONS might also usefully evaluate whether returning forms by post, rather than through enumerators, had any impact on the response rate to particular questions. (Paragraph 53)
The enumeration strategy for 2001 involved substantially higher resource input amongst hard to count groups than had been achieved in 1991. The approach to postal returns enabled enumerator resources to be concentrated in those areas where follow-up was most necessary.
ONS will analyse carefully patterns of differential non-response in the best and worst performing areas and also the impact of postal returns on the response rate for individual questions. This analysis will inform the approach to be adopted for any future Census.
(n) We note that the first results from the 2001 Census will not be available until August 2002 and that the main results will not be available until the first half of 2003, when data on which local authority spending assessments are based will be 12 years old. We recommend that ONS review the trade-off in cost benefit terms of the Census results being available earlier for users and public resource planning against the additional cost of doing so, and publish the results of this exercise. In undertaking this work, ONS should take account of the requirements of all data users, and not just the resource allocation round for local government, which seems at present to be the primary determinant of the timetable. (Paragraph 57)
The necessary trade-off between content, quality and timing of the release of output has been one of the main themes of an extensive programme of consultation between the ONS and the wide range of user communities on the 2001 Census.
Opportunities from continued technological advance continue to present themselves, which may result in a more rapid timetable in 2011.
User preferences need to be assessed afresh for each Census, as they are often driven by the experience of the previous one. As at every Census, ONS would wish to review the balance between time of delivery and coherence of results and would review whether quickly producing either provisional counts, or a limited range of final results for key variables, was preferable to the 2001 strategy.
(o) We recognise the conflict between confidentiality and precision in outputs containing small numbers. We recommend that ONS reconsider their decisions on rounding and the minimum threshold for individual output tables in the light of concerns expressed by data users and others in the evidence submitted to the Sub-committee. (Paragraph 59)
The National Statistician has a legal obligation not to reveal information collected in confidence in the Census about individual people and households. To assure the success of any future Census, it is essential that this obligation is not only fulfilled, but also that it is perceived by the public to have been fulfilled.
The National Statistician recognises the concerns about the several effects on the output that will result from the decision on rounding and is currently considering an alternative method of rounding which will address some of the main concerns.
There will be an assessment of the technical feasibility of the new proposed approach and consultation with users before any final decision is made.
The National Statistician has reconsidered the decision to raise the thresholds for the Census Area Statistics but has concluded that the basis on which the original decision was taken has not changed.
However, in the light of comments, he has agreed that for those Civil Parishes and other administrative geographies below the threshold but containing 50 or more people and 20 or more households, a set of summary statistics similar to those available in the 1991 Census Ward and Civil Parish Monitors will be produced.
For areas with less than 50 people or 20 households, only (rounded) counts of the total number of residents and resident households will be released.