Memorandum by the Royal Statistical Society
1. The Royal Statistical Society warmly
welcomes the Treasury Sub-Committee's Inquiry into the 2001 Census
in England and Wales. However, it is too early to be able to provide
informed comment on all but the preparatory work and consultation.
Therefore, the Society has decided that part of its evidence should
be aimed at providing the Sub-Committee with relevant questions
to ask a variety of parties, as the analysis of the returns proceeds
and the results become more clear.
2. The Census Offices conducted extensive
consultation over an extended time period with all interest groups.
The Royal Statistical Society was very pleased to be involved.
The Census Offices went to great pains to try to meet user requirements.
This extended not only to the content and wording of the questions,
but also to the methodology. They should be congratulated for
3. By working with the research community
and introducing the One Number Census, the ONS has taken a lead
in methodological issues. It should be congratulated for this.
4. The Census Access Project and the decision
to publish most outputs on the web will make the Census much more
accessible than ever before. This accessibility is hugely beneficial
and represents good value for money.
5. The Royal Statistical Society urges the
Treasury Sub-Committee, if it has not already done so, to call
for evidence in the near future from the company which is doing
the bulk of work of scanning the UK Census formsLockheed
Martin Mission Systemsso that it might learn how well that
process is going.
6. The Royal Statistical Society hopes that,
in the coming months, the Treasury Sub-Committee seeks answers
to the following questions:
When will the ONS present the results
of the Census?
Are there plans to have the option
of a web-based census form for future censuses?
If the results show that deprived
areas had the lowest enumeration rates, will future censuses devote
more resources to those areas?
It is too early to say, but anecdotal
suggests that responses were relatively hard to gain from locked
blocks of flats, student housing, and single-person households.
As there are many more than ever before in each of these categories,
how does the ONS propose to obviate this problem in the future?
When will the quality of the Census
be revealed? In particular:
What is the ONS estimate of the number
of addresses that never received a Census form in the first place?
How is that estimated?
What action is being taken when an
incomplete or a poorly completed form is found in the processing?
What information is available about
the response rate in local areas? What was the lowest enumeration
What is the assessment of the impact
of illiteracy on returns?
On the basis of those returns analysed
to date, how many people chose not to answer each question?
7. The machinery of census-taking has, in
the past, been through a cycle of winding down and then gearing
up again as the next census nears. Instead there should be a continuous
programme of review, discussion and improvement. The Royal Statistical
Society looks forward to being able to help in such a continuous
programme by providing expertise and a forum for discussion.
9 November 2001
2 Journal of the Royal Statistical Society Series A
(Statistics in Society) (2001), 164, Part 3, pp 423-425. Back