The pending official revisions to population estimates, the nature of
which we don't know, don't quite add up to 'The end of the One Number
Census Idea' (Ray Thomas, 5 Sept). The revisions are due on 19
September, and given the lack of firm evidence for the current
population estimate I expect them to be corrections to calculations
rather than a new figure that everyone will applaud.

The only current ONS investigations into the quality of the census-based
populations are looking at the census address lists for Manchester and
Westminster; these have not yet reached a point to affect the population
estimates.

In whatever way the population estimates do get changed, it has few
implications for the One Number Census idea, which is about estimation
of those missed from the enumeration. This is standard practice in
surveys, usually through strata weights to avoid bias in the published
results. Only its application to the Census is new.

It is important to respond to the uncertainty in population estimates.
As Ray says, by understanding the reasons for non-response, including
its reflection of the relationship between government, statistics and
the people. There is also a danger of the official population becoming
defined as those who register or respond to government enquiry,
disenfranchising and criminalising all who don't, and firmly attaching
population statistics to the administration of government. A
government's feasibility study of a national administrative population
register has been completed, currently limited to the idea of making
administration more efficient - but ID cards and citizenship are around
the corner.

Ludi

******************************************************
Please note that if you press the 'Reply' button your
message will go only to the sender of this message.
If you want to reply to the whole list, use your mailer's
'Reply-to-All' button to send your message automatically
to [log in to unmask]
*******************************************************