Select Committee on Treasury Minutes of Evidence

Memorandum from the Statistics Commission


  1.  This paper describes how the Statistics Commission fits into the new National Statistics arrangements. It goes on to describe how the Commission is approaching its task, both in terms of the organisation/ways of working and of the strategic approach, and to set out the emerging shape of its work programme.


National Statistics' framework

  2.  Framework for National Statistics, published in June 2000, set out the role and responsibilities of the Statistics Commission:

    "2.2  The Statistics Commission is independent of both Ministers and the producers of National Statistics. It has its own budget and is able to commission its own activities; the Chancellor, as Minister for National Statistics appoints the chairman and members in accordance with the Code of Practice published by the Commissioner for Public Appointments. The Commission produces an annual report, to be laid before Parliament by the Minister for National Statistics. It advises on the quality assurance and priority setting for National Statistics and on the procedures designed to deliver statistical integrity. It advises Ministers of areas of widespread concern about the quality of official statistics, so that Ministers can take these considerations into account in determining priorities and making decisions about National Statistics."

Status, finance and reporting arrangement for the Statistics Commission

  3.  The Commission is an independent, non-statutory, advisory non-departmental public body (NDPB). It obtains its grant-in-aid funding from HM Treasury, although it does not have crown status. (Grant-in-aid funding confers a degree of financial and thus operational independence beyond that normally enjoyed by advisory NDPBs.) It is independent of both ministers and the producers of National Statistics, and is supported by a small secretariat.

  4.  Framework for National Statistics requires the Commission to "review the need for statistical legislation after two years and report back to the Minister for National Statistics, and keep the legislative framework under review thereafter". This review will cover the wider legislative framework within which ONS and National Statistics currently operate, as well as the status of the Commission itself.

  5.  A financial memorandum and management statement, together with the National Statistics framework, set out the broad framework within which the Commission will operate. These were initially drawn up by the Treasury, in its role as sponsor department, and are currently being reviewed in consultation with the Commission. The National Audit Office will audit the Commission's annual accounts, and plans to conduct a dry-run audit for this year's accounts to the end of December 2000 early in 2001.

  6.  The Commission will produce an annual report to be laid before Parliament by the Minister for National Statistics. This will comment on the annual report of the National Statistician and on the way the Commission has fulfilled its remit. Subject to final decisions on timing, bearing in mind this requirement to comment on the National Statistician's report on its own, the Commission intends to publish a report and accounts in respect of its first (part) year of existence, ie the financial year to 31 March 2001 in summer 2001.


  7.  As a new body we are still developing our organisation and ways of working but this section sets out our current approach and how we see that evolving in the near future.


  8.  Current budget is £1.35 million. In the first year over half of this is being spent on set up costs relating mainly to accommodation but also to internal IT systems and setting up our website. In 2001-02, our first full year, the balance will change with staffing and related costs and externally commissioned research and audit work likely to account for the great majority of spend.

The Commissioners

  9.  The Chairman and the seven commissioners are part time. The Commission's formal meetings have a key role in giving a corporate and strategic steer to the work but after the initial period they will be relatively infrequent as we do not see them as the only or indeed the main mechanism by which individual commissioners will contribute their experience and expertise to taking the business forward.

  10.  So far the Commission has met formally twice (in September and October) and plans to meet at end November and subsequently every other month. This has been supplemented by two informal meetings of commissioners over the summer to share initial ideas and information. Other activities undertaken or in train include:

    —  a variety of representational and information exchanging meetings including the chairman addressing the Royal Statistical Society and attending a pre-Monetary Policy Committee briefing;

    —  meeting the National Statistician (who attended part of the October meeting) to discuss his plans;

    —  examining a National Statistics quality review report to help inform discussions about how we should deal with these in future;

    —  visits to main ONS data collection sites to improve our understanding of how National Statistics are produced;

    —  face to face meetings and phone discussions with the secretariat to help develop work programme proposals.

The staff

  11.  Currently the Commission has a staff of four with arrangements in hand to recruit a fifth. Although our budget would allow for recruitment of slightly more staff we have decided that at least initially, and probably in the longer term, we would prefer to retain some of this budget to use on commissioning research and audit projects from outside experts rather than trying to recruit in a small audit team expertise in all the areas we might wish to examine. This should also allow us to make use of recognized experts in any particular field who are more likely to be available for specific projects. The five posts are as follows:

    —  Chief Executive: Accounting Officer and responsible for overall management of the team including development of the work programme. Currently taking the lead role within the team in developing links with stakeholders.

    —  PA to the Chief Executive: the post holder also acts as secretary to the Chairman on Commission business liaising with his other office and will become "Webmaster" once development work on our new website is completed.

    —  Head of Secretariat and Resources: Secretary to the Commission and currently mainly involved in activities relating to the set up of the office. Once this has been completed the post holder is likely to take a lead role in supporting the Commission in its consideration of broader and less technical issues such as reviewing the need for statistical legsilation.

    —  Statistical Audit and Research Manager: currently being recruited and will have a dual role. We are looking for someone who can commission and manage audit and research projects but who can also undertake some statistical audit and scoping work personally. This flexibility will be important as to allow us to achieve the appropriate balance between internal and external conduct of our work.

    —  Secretariat Administrator: currently provides support to Head of Secretariat and also works with the Chief Executive's PA on general office administration and enquiries. This post is currently filled on a temporary basis as our long-term needs are not yet clear.

  The first three posts were filled by secondment from Government departments but the Statistical Audit and Research Manager is being recruited by open competition.

Transparency and openness

  12.  Framework for National Statistics says that the Commission "will operate in a transparent way with the minutes of its meetings, correspondence and evidence it receives, and advice it gives, all normally being publicly available for scrutiny." The Commission is taking a proactive approach here. All Commission meeting papers and minutes will be placed on the website once it has been extended to facilitate this and in the meantime we are offering to e-mail the minutes on request. We are also preparing an overall openness policy and considering the best means of allowing easy access to correspondence and other material.


  13.  Our core task is to advise on the quality, quality assurance and priority setting for National Statistics, and on the procedures designed to deliver statistical integrity, to help ensure National Statistics are trustworthy and responsive to public needs. While this remit is in many ways an indivisible one we have found it helpful in considering how we should discharge it to identify three broadly distinct but associated characteristics of National Statistics:

    —  integrity;

    —  quality;

    —  relevance/user responsiveness.

  These are discussed below and will be constant themes running through our work programme for 2000-01 and future years.

  14.  In preparing our work programme we are aware of a very large number of issues to be tackled and the need to prioritise. Our first objectives are outlined in paragraphs 21-29 below. Contributions from the Commission must be thoroughly researched and considered if they are to be effective and we will need to resist responding without due study and consideration to topical issues just because they are topical. We see our role in all three areas as checking that National Statistics has proper processes in place and that they are delivering the required outcomes. We cannot and should not do their quality assurance or user consultation for them and any perceptions that we were doing so would risk undermining their work. At the same time we have been conscious of the need to take account of stakeholders' perspectives and also to move quickly onto substantive work. Projects need to be of value in their own right but we will also be looking to draw general points from specific pieces of work.

  15.  We will take this programme forward in a way which makes best use of the fact that we are a small but flexible organisation, relying on others to do most of the detailed work and basing our strength ultimately on the considerable powers of openness.


  16.  We have used the term integrity in a fairly narrow sense of independence and freedom from interference. To some extent our role here is essentially keeping and being seen to keep a watchful eye and responding if and when things go wrong. We believe however it is important that we should be seen to be alert in this area and with this in mind we have explicitly asked the National Statistician to notify us should he be directed by a minister on a professional matter.

  17.  We are very much aware of the opportunity which will be presented by the forthcoming consultation on the new National Statistics Code of Practice. We will be looking carefully at the whole new Code and taking a particular interest in the integrity and release practice aspects.

  18.  We are also conscious that perceptions of integrity play a key part in determining public confidence in National Statistics and we will be addressing this latter issue in the context of our work to build links with stakeholders.


  19.  In looking at the quality of National Statistics we are concerned with fitness for purpose. It is important to get the numbers right (as well as collecting the right numbers) but no set of statistics is likely to be 100 per cent accurate. What matters is that the data should be good enough to meet users main needs and that their limitations should be understood and communicated. Quality issues will be the main focus of the "audit" and related projects in our work programme although some projects will have a user responsiveness or communications aspect too. There are also two related generic issues which will make important calls on our attention in the coming months. As the new procedures for National Statistics Quality Reviews are developed we will need to address two questions:

    —  How should we influence the priorities for these reviews?

    —  How can we comment on the reports of the completed reviews in a way which will add real value to the process?

  We do not know the answer to either of these questions yet but will be giving high priority to addressing them over the coming months.

Relevance/User responsiveness

  20.  Our role on this third aspect involves checking that National Statistics takes proper account of the full range of its users' needs. While we cannot and should not conduct the National Statistician's detailed consultations for him we regard this as a key area on which we need to be well informed. Some of our proposed audit projects (for example the one concerned with the adequacy of National Statistics to monitor the Government's National Cancer Plan) include an element of this and we will be looking at the developing National Statistics consultation process itself, but the main focus of our work in the user areas this year will be as a key part of a wider project to develop links with identified National Statistics stakeholders.


  21.  Although much of our work is defined by the developmental and general approach described earlier we will also be setting in hand some more specific projects. Preliminary work on these will be undertaken by the Commission itself and its secretariat, seeking appropriate factual input from National Statistics, but we will explore various ways of working and intend to outsource some of the work to independent experts (such as academic methodologists or subject specialists) in due course.

  22.  Our initial programme will not attempt to set out a firm timetable for this work but rather to indicate what we plan to do and where appropriate in what order. We intend to initiate work on all the projects discussed below within the financial year 2000-01 but timetables for completion will depend on the outcome of initial scoping work and on decisions on relative priorities made as our plans develop.


  23.  In our first year or so of operation however, we also need to concern ourselves with the infrastructure, both physical and conceptual, in which we are working. Issues here range from the simple practical ones of setting up and equipping the Commission's offices to ensuring we set in hand any necessary preparatory fact finding to support our specific remit to advise on the need for statistical legislation in due course.

"Audit" and related projects

  24.  As well as commenting on the National Statistics processes we also have a remit to carry out spot checks and if necessary carry out or commission our own audits. We have identified two areas to which we are giving priority to examining initially:

    —  the extent to which National Statistics assesses and communicates information about the reliability of its figures;

    —  the adequacy of National Statistics to monitor the Government's National Cancer Plan.

  We see these projects as having value in their own right but also as being complementary to each other and having the potential to allow us to draw out general points. We will be considering the rest of our work programme at the next meeting and strong candidates for a second tranche of work are:

    —  a review of a methodological topic such as the use of seasonal adjustment;

    —  a review of the quality of regional GDP estimates;

    —  a check on progress with implmentation of the Review of the Revision of the Average Earnings Index.

  We will however continue to plan flexibly in the early stages to allow for our priorities to be refined in the light of our developing links with users and other stakeholders.

Building links with stakeholders

  25.  The White Paper Building trust in statistics envisaged that the government would ask the Commission to "recommend to Ministers . . . its own machinery for covering the interests of users and producers of National Statistics . . .". While we have not received such a formal request we have taken up the spirit of this and extended our interpretation of it slightly to cover all three main groups of stakeholders: users of National Statistics, producers of National Statistics and providers on raw data on which National Statistics are based. We believe that we need to understand the perspectives of all three groups in order to fulfil our remit properly and we recognise that these three broad groupings are far from homogeneous (this point is discussed further as regards users below) and there are some individuals and organisations which belong to more than one of the classes. For example individual businesses and local authorities may be both users and providers and some statisticians within government who are both users and producers. There are also a number of key stakeholders, ranging from professional bodies such as the Royal Statistical Society to Parliament itself, who combine a user role with a broader one.

  26.  Our general approach to this area is less structured and more informal than might be implied by the term "machinery". We are open minded about the appropriate arrangements in the long term but felt it was right initially to build informal links in a flexible way. The rest of this section discuss our approach to building links with users where we have mainly focused our efforts so far but we plan to take a similar approach to building links with producers and providers and have already established some links with producers.

  27.  We are conscious that some good links between users and producers of National Statistics exist and that work is in hand following Framework for National Statistics to develop these further. We are seeking more information about these which form part of the context within which we need to develop our own links for two main purposes:

    —  understand the perspectives of these key stakeholders;

    —  more specifically to allow us the check on or "audit" whether National Statistics consultation arrangements are working properly.

  We are not trying to duplicate National Statistics consultation arrangements—as well as this being inefficient there would also be a risk that confusion about roles might undermine effective consultation.

  28.  There are also a number of well-established statistics user groups dealing with different areas such as business statistics or health statistics brought together under the Statistics User Council (SUC) umbrella and we have seen establishing links with these as important.

  29.  Users of National Statistics are however more diverse than reference solely to these expert users outside government would imply. It is still an oversimplification (for example not addressing the particular interests of the substantial number of local government users of National Statistics) to indentify three main classes of users:

    —  users within government;

    —  expert and specialist users outside government;

    —  general users: members of the public whether as voters or as users of public services.

  Because of the diversity of the user community we will need correspondingly diverse and flexible approaches but have not yet formed a view on wheter we should be aiming in due course for a single formal mechanism supplemented by informal links or a multiplicity of communication channels. The SUC conference on 9 November (at which the Chairman will be speaking) provides an opportunity to take further soundings of expert users but we are also conscious that we have more work to do over the next few months to develop our understanding both of current user concerns and of how we can sensibly tap into and supplement existing structures—which are themselves changing. Formal consultation about longer-term arrangements may be appropriate. One aspect we know we need to address but with which we have not yet made much progress, is the needs of general users who are unlikely to participate in user groups and an issue closely related to this—that of assessing public confidence in National Statistics.


  30.  The Statistics Commission has established a physical and practical infrastructure and the commissioners' early meetings have clarified our strategic approach and sense of direction. In the coming months increasing our visibility and consolidating the work we have already begun on building links with stakeholders will be vital but we will also to make a start with substantive "audit" projects as well as commenting on the draft National Statistics Code of Practice. In doing so we will need to be responsive to the views of stakeholders and to be prepared to evolve flexibly as we better understand their needs.

October 2000

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2001
Prepared 18 January 2001