Examination of Witnesses (Questions 60-79)
MS KAREN
DUNNELL, MS
JIL MATHESON
AND MR
IAN COPE
21 FEBRUARY 2007
Q60 Mr Todd: Yes.
Ms Dunnell: Yes.
Mr Todd: We may wish to test that.
Q61 Mr Newmark: My question is less
concerned with migration, which may be the movement of people
within the country. It is to do with immigration. There are, allegedly,
over a million people floating around this country and we do not
know who they are, what they are, where they are. It is a problem
that has clearly been exacerbated over the last 10 years. How
do you propose dealing with that and identifying those people
and understanding the real size of particularly the illegal immigration
problem that we have in this country?
Ms Dunnell: On illegal, it is
very, very difficult. We cannot ask on the census form, having
identified that you might be a migrant, whether you are illegal
or not. We cannot do that. But we are changing the whole basis
of the census, to ask about people who are usually resident at
every address and those who are there just on census night, asking
people when they arrive in the country and where they have come
from and so on. That, we hope, will help to give us a snap-shot
of who is here at any one time. Indeed, the interest in migrants
has led us to think about different definitions of who is resident,
because the United Nations' definition that we traditionally use
means that somebody has either to have been here for 12 months
or to be planning to be here for 12 months, and we know that people,
particularly from countries close to us, may be commuting/going
back and forth or coming for short periods of time. We have questions
both on the census and on our large surveys which are attempting
to measure this. In fact, we have just published some findings
from the labour-force survey about the numbers of people from
other countries who are in the labour market at the moment.
Q62 Mr Newmark: By definition, as
you have quite rightly said, it is difficult to identify these
numbers, yet compared to 10 or 20 years ago, when the number was
probably far less significant, they seem to be far more significant
now and clearly no illegal immigrant is going to fill in a form
to say, "Yes, I'm here," for the concerns to which Sally
and I alluded with our experience in Southwark 15 years ago. I
am just a bit curious. Is there another way of approaching this
problem? Because it is a concern amongst the public out there.
They want to know what the numbers are.
Ms Dunnell: Yes, we are attempting
in the census to count absolutely everybody who is here. We have
explained, I hope, today all the efforts we are going to go to
in order to do that. We will use other information that local
authorities have to quality assure and to check against that.
The evidence we have from our labour-force survey, indeed, is
that people are not happy to answer the door to a labour-force
survey interviewer and submit to a very lengthy interview about
their labour-market and other kinds of experiences. Obviously
there will be a problem with people who are illegal but I think
we also know that quite a lot of illegal migrants have been here
a very long time and it may be that they are already responding
to our surveys and may well respond to the census.
Mr Newmark: Thank you.
Q63 Mr Todd: You have toyed with
having income questions before and in 2001 it was decided not
to have that. You have put income questions in the test questionnaire
which we have. Why have you changed your mind?
Ms Dunnell: There is intense interest
in having an income question from all over government, local government,
other users—mainly, of course, for being able to classify
people by their income and to identify particularly poorer groups
in the population. Because it was felt last time that this was
unacceptable—as Ian explained, one of the key research questions
in our test is to see whether or not it has an impact on response—if
it has an impact on response then we will have to think about
it very, very carefully because our priority is to get a high
response.
Q64 Mr Todd: Where do you set your
market on that impact on the response? One can imagine that there
will be some people who will decline either to answer those questions
or the whole questionnaire.
Ms Dunnell: They might do. We
have an open mind about that.
Q65 Mr Todd: Where do you set your
measurements?
Ms Dunnell: Frankly, if it has
any significant impact on response, we will think very carefully
about including it, particularly since there is incredible pressure
on the form. As you have already pointed out, the form is quite
complicated and we do not really want to make it any more complicated.
Q66 Mr Todd: To clarify that, you
aim for a 95% response, do you not? If this dropped below, say,
90% or—
Ms Dunnell: I do not think we
would want to go that low.
Q67 Mr Todd: That is very clear.
Thank you. We have had a discussion about the Jedi faith. Identity
and ethnicity are also areas of controversy in which there has
been a significant amount of lobbying. How are you intending to
address this in 2011? I should mention that I have a significant
Sikh minority in my own constituency and they have strong views
on their recognition as a separate group.
Ms Dunnell: We are, of course,
receiving letters and are deeply involved already in consultation
with many ethnic minority groups and their representatives, including
the Sikhs. Ian, do you want to say something about how we will
have to make some final decisions at some point?
Mr Cope: We launched a formal
consultation around topics for the 2011 census in May 2005, around
what topics people wanted in the census, and we published our
response to that in March 2006. In the test, we are testing a
national identity question that allows people to say they are
English, Welsh, Scottish or other, as well as the ethnic minority
question. It is largely the same as the 2001 question, although
it does add a couple of extra tick boxes for gypsies, Irish travellers
and Arabs. We are also at the moment carrying out a consultation
on ethnicity, identity and religion, largely geared to users,
around what kind of information they want and asking whether it
should be a multiple tick box, so they could tick more than one
tick box. In terms of the religion question, the Sikh religion,
the issue—and I have certainly seen conflicting views from
the Sikh community on this—is whether Sikhs are an ethnic
minority as well and that is unresolved. There is an issue around
consistency of views in the communities and then how much space
is available. If there are large numbers in a community, it makes
sense to have a tick box; otherwise there will be a writing box
that says "Other" and people can write in that space.
Q68 Mr Todd: I do not think your
formula, as defined in these questions, will satisfy the Sikh
representatives I know of. It confines the question to an aspect
of religion—and it is voluntary, incidentally, in your questionnaire.
Ms Dunnell: That is why the results
of the consultation and the results of more extensive discussions
with communities... That is not the final version. Because we
have launched the consultation, we need to wait until that comes
in. This is another area where it is very difficult. We ask the
questions, but it is how they are presented when you analyse them.
We have done a whole report, which we have published, called Focus
on Ethnicity and Religion, which uses those variables in combination
and produces a very long list of different groups. It is really,
really interesting. You could not put that on a form but you can
create it. That is the thing which is quite complicated to get
across to groups sometimes.
Q69 Mr Todd: When does that consultation
close?
Mr Cope: It closes at the end
of March.
Q70 Mr Todd: You then have a section
on language. Has there been a demand for that particular element
in the questionnaire?
Ms Dunnell: Yes, indeed; anything
that gets on to the questionnaire is in great demand. This is
from several points of view, of course. The Welsh, for example,
are very keen to understand how many people in Wales speak Welsh,
but of course we are also very interested in the number of people
coming into the country who do or do not speak English. There
are many different reasons for being interested in that.
Q71 Mr Todd: That is presumably the
reason for the particular phrasing of this question, which does
not merely ask what is their normal language, the one they speak
at home, but actually invites them to list other languages with
which they are familiar?
Ms Dunnell: Exactly.
Q72 Mr Love: One of the major problems
in the last census was a robust national address register. Do
you feel more confident that you will have that this time? Do
you have the power that will make that available to you?
Ms Dunnell: We are working on
the basis that in 2011 there will not be a single address register.
At the moment there are two main parties that produce address
registers. The Government has worked hard to bring that together
into a single address register. To date, that has not worked,
and so the approach we are using, which we are trying out of course
in the test, is to take one of them and to check it against the
other and to amplify it with data from local authorities and from
walking the streets. Effectively, we are having to produce one
for ourselves, but of course that is not the same as having a
long-term, sustainable address register. We are having to do it
for the 2011 census.
Q73 Mr Love: You talked about the
five census test areas. It says in the documentation with which
we were provided that information from that address checking has
been analysed and results will be known in January. What are the
results?
Mr Cope: We have checked the detail.
We are about one month behind what we were anticipating on that
because we had to focus on preparing for the test. We checked
100,000 households. On the ground, we identified an additional
9,000, most of which were in Camden and most of which were multiple
occupancy, so the address was on the other list but the building
had been converted into flats and those four flats were not on
the list.
Q74 Mr Love: Taking all of what you
have said now, in the material that has been circulated you are
suggesting that you will use a number of different ways to get
responses to the census form. It talks about doorstep collection,
postbag, telephone, Internet and enumerators. Taking into account
what you have said about registering addresses, are you going
to be using more enumerators or fewer enumerators?
Ms Dunnell: We are going to be
using fewer enumerators because we are anticipating, and this
of course is a question for the test, that in many areas post-out
will be very effective and that we will therefore be concentrating
the enumerators in the hard-to-count areas. We are also trying
to recruit a rather different kind of enumerator, to have more
highly skilled ones and to make more effort to get enumerators
who match the communities that they will be enumerating than we
did last time. We are hoping that by recruiting rather fewer of
them that will be more successful and we will have quality rather
than quantity and they will be very focused in those hard-to-count
areas, and of course then focusing on non-response to the postal
approach.
Q75 Mr Love: Does that mean then
that there will be more enumerators per person to be enumerated
in the hard-to-reach areas than there were in 2001?
Ms Dunnell: I think that is highly
likely, yes.
Q76 Mr Love: I hope that some members
of staff in ONS are looking at the experience of electoral registration
which is pertinent here. You mentioned earlier about people generally
being quite happy to answer questions on the doorstep. I can assure
you that is not the case in many of the hard-to-reach areas in
my constituency, and I am sure others are in a similar position.
It will be a difficult process. Yes, you are absolutely right
to concentrate on people who reflect the community and who know
the community, but are you sure that they will have the relevant
skills to be able to do this job effectively?
Mr Cope: Our current plan is not
to have 71,000 enumerators as we had in 2001 but more like 45,000
to 50,000 but focusing on the follow-up activity. We will employ
fewer people but for longer and target this just on the follow-up
activity. We are also looking to pay them more. We had big problems
with turnover, particularly in inner city areas. We recruited
people, trained them and they started the job but did not like
it and left. That turnover is very disruptive to the follow-up.
We are also testing this approach of using outsourced agencies
to recruit because we have great difficulty recruiting, particularly
in inner city areas. Agencies have staff across the country and
they use them for their core business, whereas recruiting a very
large, temporary field force once every 10 years is not really
part of their ongoing business. That is part of what we are trying
to do. The aim is to look for better skilled people and to have
them for a longer period so that they can be more effectively
trained and focused on the follow-up activity. Essentially, that
is the approach that we are testing in the 2007 test. We will
then be evaluating whether that works in practice.
Q77 Mr Love: Interestingly, people
who are prepared to do electoral registration in the leafy, middle
class areas do not seem as prepared to do it in other areas that
are more hard to reach. This issue will come back at you. At the
last census, 2% of people did not fill out any form. My understanding
is that there were 86 cases referred to the Solicitor's Office.
Do you know how many successful prosecutions arose from that?
Ms Dunnell: They were all successful.
Q78 Mr Love: They were all prosecuted?
Ms Dunnell: Yes.
Q79 Mr Love: That is something because
they never prosecute anyone for electoral registration failure.
I have asked this question numerous times. You are doing better
than they are but are you doing well enough because 2% of the
population is considerably greater than 86? Do you think it is
time for us to be a little more robust with those who do not perform
what is a legal duty for them?
Ms Dunnell: Of course, prosecution
is the last resort and is very resource intensive. One of the
reasons why so few cases were brought last time was because we
did have difficulties in the managing of information and so we
were not quick enough in some cases to establish where somebody
had absolutely refused to fill in a form. All our management information
systems that we are setting up this time will make that much easier.
Also, of course, last time we had the foot-and-mouth epidemic
and all kinds of things, which made it quite complicated at the
last minute. This is not a priority at the moment but we will
be producing a strategy for dealing with wilful non-response.
At the end of the day, yes, that will include prosecutions.
| 
