I am delighted to have been given the opportunity to speak to you today at this most important conference in the statistical calendar. I am not, as you know, myself a statistician, but I have used statistics all of my working life and I have worked closely with producers of statistics in three government Departments. I am very proud to have been recently elected as an RSS Fellow.

I see the relationship between the Royal Statistical Society and the UK Statistics Authority as having the potential to develop into an alliance of real substance and purpose. While our perspectives will sometimes be different, our goals in relation to official statistics are, I believe, very close.

At heart, we want to see the statistical service, provided to society as a whole, steadily developed in the public interest. That is my statutory duty as well as being, I believe, our shared aspiration.

Of course, the RSS is the ‘senior’ family member at the table. It has been going strong for the best part of two centuries while the Government Statistical Service has just celebrated its 40th birthday. The UK Statistics Authority is an infant of only five months.

Now, I wouldn’t want to take this analogy too far but we could say that the Authority needs to learn to walk before it can run – and it is experiencing a few teething problems. Bear with us and watch us grow.

This is not, in fact, the first time I have spoken at the RSS. In April, I was invited to speak to the Society’s Official Statistics Section, just a few days after the Authority assumed its statutory role.

There were some very acute questions posed from the floor at that meeting – perhaps rather more fully developed than my answers could be at such an early point in our existence?

Today, we have another opportunity to exchange views, and I would encourage you both to raise questions directly with me and to participate in the separate panel discussion which will follow this plenary session.

David Lipsey, John Pullinger, and Ian Maclean are among the very best informed commentators on the world of official statistics and I am sure this will be a most stimulating debate.

I understand that the focus in that session will consider what the measures of success for the Authority should be. Statistics Authority staff will be taking careful notes of points made and the full Authority Board will discuss the messages arising from it at its meeting later this month.

Of course, the RSS is not just a world-leading professional statistical association, but it also brings together an immense fund of expertise in the field of UK official statistics including the crucially important user perspective in the form of the Statistics Users Forum.

The merger of the Statistics Users Forum into the RSS structure offers a ready-made platform for dialogue about the needs and views of the user community, and we are keen to find the best way to build on that in the months to come.

The UK’s statistical system lives, as we all do, in a fast-changing world.

Developments in information technology bring new opportunities and expectations to the table, but they also raise the expectations of the user.

Increasing population mobility makes counting people accurately much more of a challenge for our statisticians and interviewers.

Declining trust in authority and an increasing perception of a surveillance society makes statistical work much more difficult, in particular securing high response rates to surveys and form-based data collection.

The statistical system has to react to the changing public expectations of public services. New ways of working, for example those brought about by devolution and the added emphasis on target-based delivery in the public sector, means that the framework in which official statistics are produced and used, and indeed interpreted by the wider public, has changed dramatically in just a few years.

Since we live in such a fast-changing world, I feel it is vital that we have a coherent planning system for UK official statistics. Instead of a “scatter-gun” approach, composed of numerous individual departmental plans, we need a more structured and informed method of planning across the statistical system.

It is the Authority’s role, I believe, to sponsor and encourage statistical planning with user engagement at its core. In particular, the user community can help us identify issues currently in the statistical system that present real-life challenges which are not already taken account of, as well as providing expert input into gathering an evidence-base of user requirements.

I would like to say a little about what we have been doing during our formative months. We have begun in earnest, turning into action the formal remit given to us in the Statistics and Registration Service Act – “to promote and safeguard the production and publication of official statistics that serve the public good”.

In particular, I would like to touch on four things: ƒ

  • How we have set the Authority up in order properly to discharge the responsibilities given to us. ƒ
  • How our work is overseen by Parliament and how such scrutiny arrangements secure our own independence from the government of the day. ƒ
  • How we are developing our own scrutiny mechanisms for official statistics through the Authority’s assessment and monitoring process. ƒ
  • And, how users and experts will play a vital role in ensuring the work of the Authority is “grounded”, focused on priorities, and targeted in the areas that matter.

Setting up the Authority

The Authority has two distinct roles set out in the legislation.

First is what we might call ‘scrutiny’ of the entire UK statistical system. We have a formal role to raise any concerns we have with statistical work in any of the dozens of organisations responsible for official statistics – I am being careful here not to put a figure on the number of these organisations as we might then have to announce large revisions.

Also, under the “scrutiny” heading, we have the formal assessment of areas of statistical work against a Code of Practice which the Authority itself will determine.

The broader scrutiny role will lead to a series of reports we are calling “Monitoring Reports” and the assessment work will lead to reports we are calling “Assessment Reports”.

The consultation document on the Code of Practice, which closes at the end of this month, is the first of the Monitoring Reports and I will return to our plans for further reports in a moment. Do let us know what you think of our draft Code. The consultation closes on 30th September.

The second main role is our direct responsibility for the Office for National Statistics as the UK’s national statistical institute and the producer of many important statistics.

These two roles for the Authority need to be kept largely separate – not least to ensure confidence in the assessment and monitoring of the work of the ONS itself, and this means that there are a number of important actors with distinct, and distinctly important, parts to play.

On my appointment, I took the decision that two of the Authority’s non-executive members should serve as Deputy Chairs, one as Deputy Chair with responsibility for oversight of the UK statistical system, and the second with responsibility for “all-things ONS”.

Lord Rowe-Beddoe – David Rowe-Beddoe – is our Deputy Chair with responsibility for oversight of the Office for National Statistics. His background is a distinguished career in business, in the UK and internationally, and also in heading public sector organisations. David chairs the ONS Board which now meets regularly to monitor and scrutinise the work of the ONS. The ONS Board is, in effect, a sub-committee of the Authority and raises ONS issues at Authority meetings where this is required.

Adrian Smith – familiar to many of you as a distinguished former President of the RSS – was appointed as the Authority’s other Deputy Chair with responsibility for oversight of the UK official statistics system.

Adrian has chaired the second of the top-level Authority committees, the Committee on Official Statistics, which is responsible for oversight of the statistical system.

However, as sometimes happens with the best-laid plans, the unforeseen has happened: Adrian has had to step down from the Authority following his appointment to a senior post in the Civil Service. That is our loss but we are actively recruiting for his successor.

I can think of no better a forum than this RSS conference to pay a personal and sincere tribute to Adrian for his hard work and advice in the formative months of the Authority’s life, for the support he has given to me, and to wish him, on behalf of all the Authority members, every success in his new appointment.

The Committee on Official Statistics will be the main arena in which the Authority engages with users and other important external voices including the RSS and user groups, helping the Authority take a view on the shape and state of the UK statistical system and identifying things that need a closer look in our Monitoring Reports.

It is still early days; the Committee has met only once and is currently without a chair – an example of one of those teething problems I mentioned just a little while earlier. Please bear with us and the Committee will soon get the wind in its sails and will be coming to you for your views, if you are not already making those known.

Let me also mention two other important actors – both well known to you – who are central to the Authority’s work, Karen Dunnell and Richard Alldritt.

Karen, as National Statistician and head of the Government Statistical Service has professional oversight of statistical matters across Government and, while also being chief executive of ONS, Karen is our most senior adviser on the activities of the wider GSS.

Karen’s standing as the professional adviser to the Authority, of which she is also a full member, is recognised formally in the legislation.

Richard Alldritt, the Authority’s Head of Assessment, is another central player in the team. His role is, again, formally recognised in the legislation. In particular, we look to him and his new team to assess compliance with the Code of Practice and lead all our monitoring work.

At the moment Richard is leading the consultation on the Code of Practice, building up a team in London, Newport and Edinburgh, and developing our plans for the first series of reports.

Richard will also be responsible for providing responsive advice to the press and others who raise matters of immediate concern with the Authority.

As this scrutiny role develops, an increasing amount of the Authority’s time is likely to be devoted to pursuing the recommendations from the various Assessment and Monitoring Reports. So we can expect to see something of an evolution in the balance of the Authority’s priorities as we mature into an established part of the statistical system.

I’d also like to remind you about the development that we call “the Publication Hub”. The Hub is a vital tool in demonstrating a professionally independent statistical service, as well as providing a “one-stop-shop” for statistical releases which can only be to the benefit of all users.

From April this year, all National Statistics have been released online on the Publication Hub, in one central place, from where users are taken straight to the statistics without any ministerial or policy commentary alongside them.

The Authority has also hosted two National Statistics press conference events in recent months, the first of their kind. The intention is that these events will provide a forum for professional statisticians to publish and introduce their statistics, make presentations to journalists, and take questions on the statistics.

No ministers are sitting alongside them, and journalists are directed back to departmental press offices for any policy or ministerial comment on the implications of the statistics. As some of you will know, we had a teething problem here recently about which your President has written to me. Suffice it to say that I am fully in agreement with David Hand, and am pursuing the matter he raised with me.

What these innovations show is the extent to which the Authority is now able to reinforce the professional independence of statisticians in government, and the independence of the statistics they produce.

The online Publication Hub is being improved all the time. The next phase of developments is due to be launched towards the end of this year. Do let us know what you think it.

Scrutiny: Assessment, Designation & Monitoring reports, and Parliament

I have said a few things about the scrutiny role already. It straddles all the component parts of the statistical value chain – from the planning and funding of statistics, through to the collection of statistics, the delivery of statistical products, and the communication and dissemination of statistical messages.

The legislation gives us “teeth” and we will, I am sure, be able to have some “bite” in terms of scrutiny of the UK statistical system.

But we will not bite without careful process, taking full account of the need to retain the confidence of those who produce the statistics. We must be deliberate, considered and consistent in our actions and in our comments.

We should use the statutory responsibilities given to us to encourage and facilitate progress, pointing to the scope for improvement rather than criticising past practice. The visible deployment of our teeth will be reserved for the more – how should I put it – recalcitrant cases.

Central to our practical authority – the teeth if you like – is our formal accountability to Parliament, and the reports we make to them.

We value deeply Parliament’s own scrutiny of us in monitoring progress against our statutory objectives, and it is right that Parliament takes a keen interest in what we do, ensuring the statistical system delivers trusted statistics that serve the “public good”.

The Authority is required under the legislation to lay its reports before Parliament and the devolved Parliaments. That, in itself, demonstrates how we are accountable to the public, and how the Authority is different to the previous arrangements.

The UK parliamentary committee that will take the most direct interest in us is the House of Commons Public Administration Committee, chaired by Tony Wright MP. Our relationship with the Committee is still at a very early stage, but I have met Tony Wright on several occasions, and I am looking forward to working with the Committee.

Before the summer recess, the Committee took evidence from representatives of the statistical user community, including senior RSS figures, to start to identify the important messages coming from users about what the priorities are for the Authority.

I hope and expect to have an opportunity at some point in the next few months to tell the Committee what the Authority has been doing and to hear the views of the parliamentary community. The Authority will also be meeting parliamentarians more informally at an event we are organising in November.

I expect we will also be called before other parliamentary committees, in Westminster, Edinburgh, Cardiff and Belfast when these committees need to hear from us about our views on particular statistical matters.

I welcome these opportunities to report to the Parliaments, and I look forward to them. It is a key part of how the new statutory system will work.

I would like to speak a little about the how the Authority sees its own scrutiny function, rather than how its work is scrutinised from above.

As many commentators have already noted, the legislation contains something of a tension between, on the one hand, executive scrutiny over the ONS, and on the other a much broader scrutiny role over the whole statistical system.

While we have to spend, quite rightly, some of our time closely involved in ONS business, we want to spend at least as much time in ensuring our independent voice is heard on matters of public concern.

We will keep our collective eye focused on the big picture, to deliver system-wide benefits and build long-term public trust in the products and outputs that make up the statistical service.

We announced our plans for the first set of Monitoring Reports in July because we want to make rapid progress in reviewing those areas that have a particular relevance or importance to the statistical user community and wider society.

In this context, I am most grateful to the RSS, especially the National Statistics Working Party chaired by Jill Leyland, for keeping us informed of what they think are the most important issues for us to consider. If our initial plans have not done full justice to the RSS views, then of course we will be glad to discuss them further.

I have already mentioned that the first Authority Monitoring Report is the consultation document on the Code of Practice. And the second, in early January I hope, will be our final report on the Code.

But we are also preparing a report on progress in improving migration statistics. While there is a lot of work currently being done within Government to improve measures of migration – and Karen Dunnell is working closely with an inter-departmental group of Ministers to deliver this – it is important that the Authority comes to a view on whether everything that can be done is being done.

Possibly before that is published, we will issue a root-and-branch review of all those official statistics that are not currently designated as National Statistics with a view to making recommendations for the extension of the set identified as National Statistics – again there are specific legal powers for us to do this.

The Authority has this power to issue notifications that statistics should be brought within the ambit of the Code and the Authority fully intends to use it. We want to make sure that the Code of Practice applies to all those statistics that professional experts think should be covered.

In the next twelve months we also plan to publish other Monitoring Reports on:

  • the communication of measurements of inflation (and consumer prices);
  • barriers to trust in relation to crime statistics; • the adequacy of environmental statistics to inform public debate and government policy;
  • And finally, and I spoke about this a little earlier, arrangements for longer term planning for statistics to meet society’s needs.

As soon as we are little closer to finalising the Code of Practice, the Assessment team will begin the task of assessing against the Code the current set of 1,300 National Statistics products, as well as the significant number of statistics that are currently in the “waiting room” for designation as National Statistics.

1,300 may not sound all that many, but it is worth remembering that the Census counts as just one of that 1,300. The complete assessment of National Statistics is a Herculean task.

Just as the Code covers all aspects of statistical planning, production and communication – ranging from the funding and planning of statistics through to the communication of statistical messages to those who need to use them – so too will Assessment look at the entire process.

In that sense it is not the statistics, as such, that are being assessed. It is the entire service being delivered to the user.

While some of our attention will focus inwards on the statistical products of government, we will also closely monitor the issues and concerns that people such as yourselves, journalists, and other interested parties, have raised with us.

We will be very open about the concerns that have been raised with us, whether we share those concerns or not.

If you look on our website this week, you will find what we call our ‘Issues Log’.

This has only just been started but it is intended to be a fully public record of concerns that have been raised with us. The log is maintained by our Secretariat and Assessment teams, who will also add issues that they have identified in the press and news media.

We will develop this in the light of reaction to it. The inclusion of an item on the Issues Log does not necessarily mean we are planning to devote resources to investigating it, but it does mean we are publicly acknowledging that we are aware of concerns.

Involving users and experts in the work of the Authority

I have said that we want to involve users and experts in our work but, so far, I have not said much about how we will do this.

As we set out in our consultation document on the Code of Practice, which also covered the principles and procedures for assessment work, we will be actively engaging users and external experts in planning and guiding our reports. We will also invite appropriate experts to sit on project boards, to tender for research contracts or to offer advice on specific issues.

All of this is still to evolve but we will make sure that it does. To the greatest extent possible, we want our reports to reflect a consensus among the expert community. That is the way to bring maximum influence to bear in those cases where influence is needed.

And, of course, we also want these reports to be grounded in a good understanding of the user requirement, both today and in the future.

This is not an exact science. But we must do all we can to focus statistical production on real current and future requirements rather than simply what is easiest, or least controversial, to produce.

We must also avoid the temptation to produce reports that are too academic or rarefied in their arguments. Our main audience consists of those who use statistics, and those politicians, journalists and commentators who influence opinion about statistics.

These people must be engaged with what we are saying, understand our message, and support us in pushing for improvement.

That may mean producing reports that bear little resemblance to a PhD thesis and are relatively short and plain. But we must still aim for consistent high quality. That is a real challenge and one in which we will need your help.

As part of this we will be looking to establish a systematic dialogue with the many user communities, in particular the Statistics User Forum, central and local government, the National Health Service, the commercial and voluntary sectors, and of course the academic users who use statistics in so many fields of research.

The mission of the Authority

I would like to finish by returning to the central role of the Authority as laid down in legislation – “to promote and safeguard the production and publication of official statistics that serve the public good”. What does this mean in practice? Let me have a go at unpacking the meaning in as simple plain English as I can manage:

  1. We must make sure that the right statistics are produced, i.e. those that help us all understand our society and our economy, and help policy-makers in their decisions.
  2. We must make sure that the highest professional standards are maintained, and;
  3. We must make sure that the statistics are communicated intelligibly, clearly and as attractively as possible to the public. In this way we will enhance trust in the statistical system, both in terms of its quality, and its political independence and impartiality.

The UK Statistics Authority is still in its infancy and we still have to acquire skills, competence and confidence. But we are making steady progress and we have, I believe, a clear idea about the direction in which we are heading.

I look forward to continuing the dialogue with you.

Thank you once again for inviting me this afternoon.