Examination of Witnesses (Questions 180-199)
MS KAREN
DUNNELL, MR
DENNIS ROBERTS
AND MR
MIKE HUGHES
7 JUNE 2006
Q180 Mr Todd: What proportion of
official statistics is classified as national statistics? Bearing
in mind your comment about administrative data, it must be petty
difficult to work out.
Ms Dunnell: Yes. About four in
five of all GSS statistics at the moment are national statistics.
If you take away the ONS ones, then it is slightly less. Of those
statistics produced by other government departments, about one
fifth are not national statistics.
Q181 Mr Todd: You have previously
told the Sub-Committee that legislation on access to administrative
data, indeed you have just referred to it earlier, would be enormously
helpful. That is not encompassed in these proposals. Should it
be?
Mr Hughes: The document does allude
to the possibility of this being considered, and this was something
which we placed a lot of emphasis on in our discussions with Treasury
colleagues, and so I think ministers will want to consider that
very carefully in the light of all the evidence and contribution
which your Committee makes. It is something which we are, for
obvious reasons, very keen to see happen. This is a unique opportunity,
with this particular piece of legislation, to perhaps get something
done on this.
Q182 Mr Todd: This is a once in a
generation opportunity to tackle issues relating to access to
public data and statistics?
Mr Hughes: Yes.
Q183 Mr Todd: Surely it should address
something more than obsessions with whether ministers are saying
things out of turn or not, should it not?
Mr Hughes: Yes, we believe that
the opportunity for statistical legislation, as you say, is not
going to happen very often. We are very grateful for this opportunity
and we are working as hard as we can to get this particular part
of it in there.
Q184 Mr Todd: This is an opportunity
to examine a much more liberal regime over public centre data
and statistics and making those more freely available to the public
and third parties to use as they think fit?
Ms Dunnell: I think one has to
be very careful with this. Obviously, we want all the statistical
information that the public and business, and so on, provide and
to be as available as possible, but one of our other roles, of
course, is to protect the confidentiality of the data, and so
I think this has to be done in quite a controlled way under very
specific arrangements.
Q185 Mr Todd: Bearing in mind a lot
of data collection relies on the co-operation of the public in
the first place.
Ms Dunnell: Yes.
Q186 Mr Todd: In some cases they
would not mind having their data back, rather than being charged
for it or placed in some exclusive framework which makes it less
easy for them to get to it?
Ms Dunnell: No, I think that is
right. In fact, one of the things that we have tried to do in
the last five years is make all the statistical data that is available
across the system easily available, so all of our developments
on the website, for example, I think, are very important. The
whole development of neighbourhood statistics has put a huge amount
of both local and national statistical information onto an easy
to access website, and, of course, we have stopped charging anybody
for access to data; so I think those are part of our very fundamental
principles.
Q187 Peter Viggers: You were present,
I think, when we heard Dr Fellegi say that the present proposals
were really tinkering?
Ms Dunnell: Yes.
Q188 Peter Viggers: A number of witnesses
have argued with us that ministers should not receive access to
national statistics before they are published. Do you think the
Statistic Office can ever really be regarded by the public as
independent if government ministers continue to enjoy pre-release
access to statistics?
Ms Dunnell: It is a very good
question and it is something, of course, that we have spent a
lot of time thinking about, but it is also a practice which, I
think our Scottish colleague explained, is very deeply embedded
in our system, and what the document is looking for and what we
are proposing is a constraining of the pre-release access, because
I think that would be much more acceptable.
Q189 Peter Viggers: Would you perhaps
elaborate on how it would be constraining?
Ms Dunnell: Mike, would you like
to do that?
Mr Hughes: I think it could be
done, as the consultation document says, probably in three particular
areas. Picking up what Ivan Fellegi said earlier, there is a whole
raft of series where you could argue there is no justification
for having pre-release access. The second area is cutting back
enormously the number of people who actually get it. Thirdly,
reducing the time, and I think associated with those measures
is a much more coherent and consistent approach to this across
government. I have great sympathy with David Rhind's remarks about
the code at the moment, because the code is an articulation of
good practice rather than saying, "You should be doing this
or that." So it is left to individual departments to decide
how they choose to interpret it. So, a much more coherent and
consistent approach to this, an approach that is regulated through
the assessment and audit process; these are all the sorts of things
that we believe could fall within the ambit of the reductions
that Karen has talked about.
Q190 Peter Viggers: Will there ever
be true objectivity and the respect that follows that if the statisticians
share the same offices with the policy advisers?
Mr Hughes: I have been in ONS
for three years at the end of my career, but I have spent my whole
lifetime as a government statistician in departments, and I was
the head of profession for the DETR and then, latterly, transport.
Where the statisticians bring real added value is in the discussion
of the policy issues to make sure that those policy issues are
founded on evidence-based policy and making sure the information
is right. I feel that is one of the great strengths of the British
system. The ability for the statistician and the head of profession
in each department to have some kind of protection through a statutory
code is a perfectly acceptable way of working, and I think if
you actually take the statisticians out of that environment, they
are going to be far less valuable and far less effective than
they are now. That is a personal opinion, but I know it is one
that many of my heads of profession colleagues share.
Q191 Peter Viggers: We have heard
from a number of witnesses who have expressed concern at the extent
of the fragmentation of the 2001 Census and the impact it has
had on the provision or comparable UK-wide statistics. In what
ways does the concordat on statistics need to be strengthened
in order to limit the extent of fragmentation?
Ms Dunnell: One of the reasons
why we have now established, as Rob Wishart said, a new concordat
between the three Registrars General for the 2011 census is exactly
because of some of the fragmentation that did happen in 2001,
and, in fact, this was something that I led with the Scottish
Registrar General in my previous job when I had more direct responsibility
for the census. The whole point of what we are trying to achieve
there is to have a UK set of outputs, which we did not have last
time. It was very difficult. We had to construct, after the event,
a UK data set. So, the whole idea is to have a consistent set
of UK outputs which will be disaggregated by the four devolved
administrations, but what that does not mean necessarily is that
all the methods and questionnaires have to be exactly the same.
As Rob says, we have to be sensitive to the different requirements
in different countries, but the whole point of the way we are
now running the census through a UK Census committee is to ensure
that at the end of the day, however the differences along the
way are, we can still meet that remit of having UK outputs as
consistent as possible, and I think that will be a major development
for some time.
Q192 Peter Viggers: The Royal Statistical
Society has suggested that the Government's proposals to fund
the census through the usual spending review are not acceptable.
Do you think that the Government's proposals on the Census will
enable the Statistics Office to deliver effective long-term planning
for this very important census?
Ms Dunnell : What I do know, I
think as a previous witness has said, is that the Census has to
be dealt with separately, because what we are hoping to get is
a sum of money and a rather longer term commitment to that sum
of money, and that does not fit the Census model. What we are
very much working towards in the Spending Review process is acceptance
by government (and this has already begun, of course) of the longer
term costs, because we are estimating costs right up to 2013 now
and what we need to get is an assurance that that proper census
will be funded. Whatever way it happens, I do not think it really
matters. What we need to know is that we are going to get that
money. Of course, the census is very much something for the whole
of government, and I think that that is recognised.
Q193 Peter Viggers: There has been
quite a lot of concern expressed about the Government's proposals
not being sufficiently definitive of the independent governing
board, and the possible problems with the board assuming executive
functions in the delivery of statistics. What role would you wish
to see the independent governing board perform in relation to
the running of the Statistics Office?
Ms Dunnell: I think that setting
up a new governing board is a major opportunity actually to really
increase the importance of statistics and the statistical profession
across the UK. I think the board's main role will be to have a
lot of strategic debate about what statistics are for in the UK,
who they are for, whether or not they are covering the right areas,
whether they are meeting our international and other requirements.
I think they will need to ask questions about the way that they
are funded, whether they are funded adequately, et cetera, et
cetera (so a very strategic point of view), and I think that that
will be a great addition to statistics.
Q194 Peter Viggers: Will you be sharing
facilities in terms of overlap of staff? Will you be providing
the board with secretarial back up? Will they have their own facilities
in this regard?
Mr Roberts: I think that is one
of the issues we have identified we need to work through. There
are, as you indicate, two models there. One is for the board to
share the same sort of secretariat as the chief executive, and
that is practised in some organisations. In other organisations,
such as the BBC, they have taken exactly the opposite approach
and decided that the governing board should be very distinct from
the chief executive. We need to work through those different models
to see what is most appropriate in these circumstances, and that
work has not been done yet. It has been identified as something
which we need to do over the next few months and something that
we will be working on.
Q195 Peter Viggers: How far down
the road are you on identifying the kind of people you would expect
to see on the independent board?
Mr Roberts: I think that is the
same. What we have identified is that we need, as the chair of
the board, someone who will be able to speak in public, to be
a public representative and just a spokesperson for the importance
of statistics, not for the statistics themselves—that is
the role of the chief executive—but someone who will stand
up for statistics, and that will be quite a large role. So, one
is looking for a significant public figure for dealing with that.
Beyond that we are looking for people who will produce a rounded
board, bringing together different expertise from across the whole
range of users and providers and those with an interest in statistics
across the UK.
Q196 Peter Viggers: Its answerability
to Parliament. I am interested in this in terms of the answerability
of the Electoral Commission to Parliament through the Speakers
Committee of which I am a member. What thought has been given
to that?
Mr Roberts: The consultation document
has, quite rightly, said that this is a matter for Parliament
itself to say what that accountability should be. It suggests
that it might report, as the ONS currently reports, through the
Treasury Select Committee as one option. I think there is another
option that one might have a separate statistics committee devoted
entirely to this, but I think this is a matter for Parliament
and, as the proposals develop, we would hope you would provide
some guidance on that.
Q197 John Thurso: John Pullinger,
in his evidence to us, made the interesting suggestion that each
of the nations should have their own national statistician to
reflect the reality of devolution. There would, therefore, be
a national statistician for Wales, for Scotland and Northern Ireland.
What title do you feel that you should have if that were to come
to pass?
Ms Dunnell: I hope it does not
come to pass. Personally I would rather stick with the title that
I have got now, because I think that "chief statistician"
is something that has been used across other government departments
and in devolved administrations and will be quite puzzling to
people, but that is something, obviously, that we have got to
think more about and discuss with colleagues.
Q198 John Thurso: If each of the
devolved nations appointed their own national statistician, not
to increase their work in any way but to simply reflect the importance
and the reality of devolution, and you continued in your existing
role, how would you best describe your own role if you were given
ultimate professional responsibility for all UK official statistics?
Ms Dunnell: I think, in a sense,
the devolved administrations, of course, have got a leader anyway,
and Rob is one of them, and the other the administrations have
them and they work very, very closely with all the other heads
of professions in all the other government departments as part
of the sort of governing body, if you like, of the Government's
statistical service, and that is a group which I chair, which
meets very regularly. Basically, that works through all having
a common agreement about what we are aiming for in terms of serving
customers, serving governments, serving international organisations,
and, of course, they all subscribe to the code of practice. So,
that is something that we talk about all the time. They are also
very, very concerned at the moment with the development of the
profession as a whole, which, as you know, is a big part of a
cross government drive to improve professional skills in all ways.
The GSS and the devolved administrations are very much part of
all that in a very joined up way, and we have many other examples,
which have developed greatly actually over the last five years,
of working together to produce coherent statistics. For example,
when we did the neighbourhood statistics project, we set up a
formal four nations group to make sure that small area statistics
were delivered in a similar way right across the UK. The Census
example we have already talked about. We have just, for example,
produced a book of UK health statistics, and earlier in the year
I announced the setting up of a new National Statistics Demography
Centre, which is bringing together all the work across government
on population estimates, projections and demography more generally.
So, what we are trying to do is gradually take on more and more
parts of the system and have a much more UK approach to them.
Q199 Chairman: Can I clarify one
point in the document. The Government envisages your office as
a non-ministerial department, and it gives examples of others,
like Ofsted, OFT or Ofgem. Each of those, of course, has a relationship
with the Minister. The minister makes the appointment, can issue
directions and so on, and I am assuming from the document, are
you, that this non-ministerial department will still come under
the ambit of the Treasury?
Mr Roberts: It is assumed we will
need a minister responsible for statistics legislation to oversee
the working relationship.
| 
