The US decennial census is enshrined in the US constitution, although in very brief form because the original interest was to determine the number of potential soldiers, as well as total population.

There have been suggestions about alternate forms that could/would improve precision, but if those improvements even suggest that one group (i e., likely block voters) would get additional representation at the expense of other group(s), then representatives of the other group usually get all animated about their 'reduced' representation.  The words in the argument can discuss any aspect - precision, "accuracy," constitutional requirement, cost, etc. - but the real issue is votes.  Doesn't matter who's votes are 'reduced.'

IMHO, the people doing the loudest huffing & puffing on this issue are not fully cognizant of the statistical & mathematical basis for the proposers' suggestions.  That the explanations are on the subtle side, or perhaps not explained well to those outside the circle of knowledge, is an added concern.

All that said, I think some proposals that assured equal or superior statistical precision across a non-homogenous population, with projected sizable reduction in implementation costs, could get a serious hearing in Congress in these present times.  It would have to include something every 10 years, to meet the Constitutional requirements, in any case.  I don't recommend even suggesting changing that document.  A lot of people will decide their ox is the one being gored.

Jay
On Jan 6, 2012, at 10:43:18 AM, John Bibby wrote:

> I'm not up with this, but have always been rather sceptical of the
> "gold standard" claims that are made for the decennial census.
> 
> A rolling series of biennial surveys (possibly 100% in sub-areas)
> could be a strong competitor. I'd like to see some sensible
> comparisons.
> 
> JOHN B
> 
> 
> 
> On 6 January 2012 14:49, Robin Rice <[log in to unmask]> wrote:
>> Hi all,
>> 
>> I'm wondering what the general consensus is, or if there is one, on the
>> question of replacement of the 2021 Census with other types of data
>> collection?
>> 
>> The omission of the long form in Canada, which I thought was so far ahead
>> being the first to allow electronic submissions and having a 5-year Census,
>> has raised a ruckus amongst data users, but I've not heard much dissent
>> about this idea to scrap this 110 year tradition in the UK for reasons of
>> 'efficiency':
>> 
>> From the consultation document:
>> 
>> in May 2010 Sir Michael Scholar, Chair of the UK Statistics Authority wrote
>> to the Minister for the Cabinet Office to say that:
>> 
>> 
>> 
>> “As a Board we have been concerned about the increasing costs and
>> difficulties of traditional Census-taking. We have therefore already
>> instructed the ONS to work urgently on the alternatives, with the intention
>> that the 2011 Census will be the last of its kind.”
>> 
>> 
>> 
>> The consultation finishes on 20th January 2012, at http://t.co/aFrG6FR9 with
>> a questionnaire about your use of Census data.
>> 
>> There is also a discussion forum started by the Royal Statistical Society
>> Statistics User Forum here: http://www.sufenews.org.uk/forum/15
>> 
>> The Guardian Data Store just did a compilation about data from 'the Iron
>> Lady's' time in office, noting that there was no ethnicity benchmark data
>> from 1981 (as the Census only asked about country of birth then).
>> 
>> Bearing that in mind - wasn't so very long ago - won't we really be much
>> more data poor without a decennial Census? We've got new migration questions
>> in 2011. If ten years is too long to wait for benchmark data is the answer
>> really to scrap it? Are the alternatives really more cost effective without
>> sacrificing geographical accuracy and comprehensiveness? Would
>> 'administrative' solutions be any less intrusive for privacy than the
>> mandated Census?
>> 
>> I'm good at asking questions but not so much at answering them. I'd love to
>> read a view from a trusted 'radstats' perspective. Do the editors of the
>> newsletter or the newly formed Population Studies group
>> (http://www.radstats.org.uk/popgroup/index.htm) have any plans to publish a
>> paper on this topic? If not, is there someone who'd be interested in
>> cobbling together a blog post for the Radstats site (probably with the
>> approval of the troika)?
>> 
>> Looking forward to any responses,
>> 
>> --
>> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>> Robin Rice
>> 

Jay Warner
Principal Scientist
Warner Consulting, Inc.
4444 North Green Bay Road
Racine, WI 53404-1216
USA

Ph:       262.634.9100
email:  [log in to unmask]
web:    www.a2q.com

The A2Q Method (tm) -- What do you want to improve today?

You may leave the list at any time by sending the command

SIGNOFF allstat

to [log in to unmask], leaving the subject line blank.