Examination of Witnesses (Questions 260-279)
MS KAREN
DUNNELL, MS
JIL MATHESON
AND SIR
MICHAEL SCHOLAR
28 JANUARY 2008
Q260 Jim Cousins: You do not publish
it, that is one thing, but do you have it whether or not you publish
it?
Ms Dunnell: No, we do not because
we are constantly changing the methods, especially at the moment,
by which we input into the population estimates. This is something
which we will take away from this meeting and think about but
it is very, very complex methodologically because the data is
coming from such a huge variety of sources and each of the sets
of data has a different level of reliability so, yes, it will
depend very much on which element of change is present in each
local authority.
Q261 Jim Cousins: Sir Michael, that
is something you might care to bear in mind for the future. I
think that is an important point, that we do not have a margin
of error. Can I bring you to this Labour Force Survey adaptation
of the migration figures. My attention was drawn, because it happens
to be a submission that the Committee has had from my particular
part of the country, I represent the City of Newcastle in the
North East of England and the North East of England is the smallest
region—there is an issue about whether the north of Ireland
is a region or a country and I am not getting into that, but the
North East is the smallest region—to the fact that the number
of contacts relevant to international migration using the Labour
Force Survey in the North East, which is a small region, is tiny.
It varies between six a year and 24 a year over the last six years
for which we have had any information. That is six cases. Wales,
which is also a small region, has similarly small numbers, but
in the case of Wales the pattern of the answers does not vary.
In the case of the North East you have a very small sample, a
tiny sample, and the results vary quite a lot from year to year.
There is a 70% variation in these six to 24 cases over six years.
The effect in the latest three year estimates, based upon these
six to 24 cases, is to reduce the population of the North East
by roughly 9,000, but because the impact of international migration
in the North East is heavily concentrated on the City of Newcastle
almost all of that regional variation hits the city of Newcastle.
6,000 of that roughly 9,000 reduction hits the City of Newcastle.
Surely that is not a sound or robust method of calculating the
real impact of migration. I am not here asking how you do it statistically
because I can understand that, what I am saying is the result
is to produce a huge margin of error.
Ms Dunnell: I think you are actually
talking about the International Passenger Survey that is based
on those small numbers. As I said at the beginning, out of the
90 million movements in and out of the country we have to identify
migrants and sample them and that is based on what we have already
accepted is not a large enough number, which is why we are now
boosting the numbers. What we are now doing with the Labour Force
Survey, given that we are reasonably confident about the total
number, the net migration in the UK, we are now using the Labour
Force Survey to apportion them between local authorities.
Q262 Jim Cousins: The information
I have just given you is precisely in the Labour Force Survey.
Ms Dunnell: Okay.
Q263 Jim Cousins: It is precisely
after you have made those changes.
Ms Dunnell: Okay.
Q264 Jim Cousins: Right?
Ms Dunnell: Right.
Q265 Jim Cousins: So you agree that
is a serious point?
Ms Dunnell: I do not know exactly
what you are referring to there. What you are saying is we have
identified six migrants in the LFS, recent migrants in the LFS.
It is very hard without looking at what you are looking at to
see exactly what it is.
Q266 Jim Cousins: In the case of
the North East, each year in the LFS there is a discovery of between
six and 24 international migrants in the LFS Survey in the North
East.
Ms Dunnell: Right.
Q267 Jim Cousins: On the basis of
that, the overall population of the North East is being reduced
by 8,500 and 6,000 of that reduction is being attributed to Newcastle.
That cannot be a sound or robust method of being so sure about
those things now, can it?
Ms Dunnell: Have you got any more
information?
Ms Matheson: I have not without
looking at the numbers.
Ms Dunnell: We will have to look
at that in detail.
Q268 Chairman: I think we will need
a note on that.
Ms Dunnell: We will come back
to you on it.[4]
Q269 Jim Cousins: Now we move on from
the difficulties of arriving at the information to how that information
gets used. We have had a number of representations from local
authorities on this point. Do you think, and I am not here asking
the statistical margins of error, which clearly are potentially
quite great in this, it is a sound method for allocating resources
through the Department of Communities and Local Government or
through the Department of Health to use information based on this
kind of data?
Ms Dunnell: That is really up
to DCLG to answer. I would say that the Office and its predecessors
have been doing population estimates for many, many years and
they are based on data that are available. At the moment we are
working very, very hard to improve them. It is not just the total
numbers in the population that are taken into account in these
formulae, as I think I explained earlier, lots of other factors
about areas are taken account of in the formulae.
Q270 Jim Cousins: Let us just consider
those two departments, DCLG and the Department of Health. The
Department for Communities and Local Government used to allocate
resources annually with a post-dated correction for new population
information as it became available and there were data changes
which were retrospective, so there was a built-in method of adjustment.
The method from this year that the Department is using is in a
three year slab which does not change, so in the year 2010-11
the Department for Communities and Local Government will be allocating
resources on the basis of the mid-year estimate of populations
of 2004 projected forward for those years with no system of retrospective
adjustment. Do you think that is a sound method for distributing
resources fairly?
Ms Dunnell: It is not up to me
to judge what decisions my colleagues in DCLG—
Q271 Jim Cousins: What margin of
error will there be by 2011 based on such methods of doing things?
Ms Dunnell: We would have to go
away and try to estimate that. 5[5]
Q272 Jim Cousins: I would be grateful.
Ms Dunnell: What they have done
this time, which they do not always do, given they have decided
to give people a three year settlement, is to use projections
which do take account of the future changes in the population.
What they have not done is wait to do that until we had the latest
population projects, but we were in contact with them about that
obviously.
Q273 Jim Cousins: Let us look at
the Department of Health then. In the current year, 2007-08, the
Department of Health is allocating resources to primary care trusts.
They have a complicated system of weighting but the raw population
data is done in this fashion: they take the 2003 population projections
for the relevant year and adjust it for the 2004 transfers across
areas using GP registration data of a completely different kind.
For the forthcoming year, 2008-09, the Department of Health has
chosen to freeze all of this data and they are simply going to
carry the same information forward to 2008-09. So the Department
of Health will be allocating resources to primary care trusts
in 2008-09 on the basis of 2003 estimates projected forward, adjusted
by the 2004 changes between areas on the basis of GP registrations.
Does that strike you as being a sound basis for arriving at a
correct impression of the population in 2008-09?
Ms Dunnell: It may not have been
the one I would have chosen but the Department of Health have
an awful lot of information about patients in their own systems
and they have decided to do it their own way. The only thing I
can do is talk to their chief statistician and see what impact
he has had on the whole matter, but their funding decisions on
allocation are not something that they regularly consult us about.
They consult us about what we think about the estimates and projections
that we produce each year.
Jim Cousins: If you were able to do that
and report it back to us, that would be very helpful to us. Thank
you. [6]
Q274 Chairman: Professor Rhind told us
that there was some concern that local authorities had a vested
interest in finding ways to increase their population count. How
do you ensure that particular local authorities, given some of
the problems last time, are not all in the game of actually inflating
their population count at the edges where they can?
Ms Dunnell: That is exactly why
we are very, very clear that when we do our annual population
estimates we actually use a method which is comparable for every
single local authority across the country and why, when we talk
about working very closely with local authorities, this is not
so that they can influence our numbers and why we are very keen
to work with them to identify new ways of using administrative
data but we are not prepared to accept a figure on this from one
local authority which may be on a very different basis from another
local authority. We will uphold the provision of estimates that
are based on the best methodology that we can do given the data
sources that are comparable across the country.
Q275 Chairman: Coming back to the
questions asked by Nick Ainger, what are the barriers to the adoption
of a system that is more based on administrative records rather
than the Census? I think the National Statistician mentioned "cultural
barriers". What are the barriers? There are legal barriers
presumably.
Sir Michael Scholar: I think there
were some important legal barriers which are potentially removed
by the 2007 Statistics and Registration Services Act. That Act
provides gateways for opening up some barriers between government
departments which until now make it difficult or impossible to
communicate administrative information from those departments
to the ONS. I think that one of the streams of work coming out
of the 2007 Act will be to identify in a list of priorities for
moving and seeking to put to Parliament a case for opening up
this or that gateway.
Q276 Chairman: So there will be an
incentive for the Board to get on and identify where data could
be shared?
Sir Michael Scholar: Absolutely.
From the evidence that you have heard this afternoon, and I have
been listening to, it is clearly of the first importance for improving
the UK's statistical system that we should make the maximum use
of administrative sources, not only from the point of view of
accuracy and coverage but, I also suggest, from the point of view
of cost-effectiveness.
Q277 Ms Keeble: Just looking at the
construction of the Board, there was a proposal that you should
have somebody from a commercial background as a non-exec. Has
that actually been achieved?
Sir Michael Scholar: Yes, indeed.
There are two people on the non-executive cadre of the Board who
have that kind of background. Lord Rowe-Beddoe of Kilgetty has
a distinguished career in business, both behind him and he is
currently involved in a number of businesses here and abroad,
and Partha Dasgupta has, until very recently, held an executive
position in a financial institution in the City of London. So
there you have two people who have current and very recent business
experience.
Q278 Ms Keeble: They only meet formally
in April as a Board, is that right? Have you met informally as
a group?
Sir Michael Scholar: No. I think
it was before you came into the room that I mentioned the first
meeting of the Board is on Saturday 2 February, this coming Saturday.
It is our first meeting of any kind and we are planning two further
meetings before April. I say they are formal meetings, they are
formal meetings of the Shadow Board because the Board does not
exist until 1 April. We are having those meetings first of all
to be able to respond to the various pieces of consultation which
the Government has launched and seeks our view on, and also to
be ready to hit the ground running on 1 April.
Q279 Ms Keeble: One of the things
you have got to respond to is the pre-release consultation with
the Government, is it not?
Sir Michael Scholar: Yes, it is.
4 Ev 283 Back
5
5 Ev 286 Back
6
6 Ev 286 Back
| 
