Memorandum submitted by Telford and Wrekin
1.1 The success of the 2001 Census undertaken
by the Office for National Statistics (ONS) was blighted by two
high profile problems:
Significantly undercounting the population
of 15 local authorities, with the most high profile example being
Westminster. The initial 2001 Census results published in September
2002 undercounted the population of Westminster by some 22,000
(11%) people. A revised population estimate was not published
by ONS until September 2004.
The disclosure control techniques
developed and applied by ONS to protect individual's anonymity
resulted in inconsistent and sometimes meaningless small area
1.2 As well as a need to ensure these problems
are not repeated by the 2011 Census, ONS must prepare to ensure
that the Census accurately counts new migrant streams, including
those from new EU accession states, into the UK and, importantly,
migrant streams within the UK.
1.3 These requirements are challenging and
how well ONS overcomes them will be a benchmark against which
the success of the next Census will be measured. The consequences
of the 2011 Census not producing accurate figures include:
Ongoing shortfall in and misallocation
of revenue support grants to local authorities.
Inability to accurately measure local
economic and health indicators, such as Gross Value Added (GVA)
per head, and ensure economic and health policy, locally, regionally
and nationally are based on robust data.
Inadequate assessment of new migrant
and ethnic communities' needs.
Failure to target capital spend at
growth areas to deliver sustainable economic and population growth.
1.4 Telford & Wrekin Council believes
that one fundamental way in which ONS can begin to prepare to
overcome these challenges is to plan the 2011 Census as a partnership
exercise with local authorities rather than a top-down government
project. Such an arrangement will allow ONS to access the considerable
wealth of knowledge local authorities have about their areas to
ensure survey resources are most effectively targeted to maximise
the Census response rate from all sections of our communities.
1.5 In adopting a partnership approach to
the Census, ONS should work much more closely with local authorities
and ensure that all the results and outputs of the 2011 Census
are fit for purpose and meet the needs of local authorities and
other key stakeholders.
2.1 The results of the 2011 Census will
provide central and local government with key data to inform the
development of policy, services and resource allocation for the
following decade. For the full value and benefit to be gained
from this Census, it is critical that the ONS preparations focus
on ensuring that the results are robust, fit for purpose and reflect
the information needs of both local and central government.
2.2 To achieve these objectives, Telford
& Wrekin Council believes ONS must undertake the 2011 Census
in partnership with local authorities and not as a top-down government
project. There are three core areas we believe ONS preparations
for the 2011 Census would benefit from partnership working with
1. Planning and undertaking the Census fieldwork.
2. Effectively measuring population and migration.
3. Ensuring all Census output adds value.
3. PLANNING AND
3.1 The 2001 Census significantly undercounted
the population of 15 local authorities. Subsequent inquiries into
why this happened have suggested that if ONS had worked in partnership
with local authorities these errors were more likely to have been
avoided. Specific problems in the 2001 Census included:
Significant differences between the
address database used by ONS to administer the Census forms and
databases held by local authorities.
Problems identifying neighbourhoods
which were more likely to produce low response rates and provide
problems of "access" for enumerators.
Where the Census results for a local
authority were significantly out of step with pre-existing ONS
data, there was no consultation or discussion with relevant local
authorities prior to finalising and publishing results.
3.2 As a result of these problems confidence
in the Census "product" was undermined. To avoid a repeat
of this for the 2011 Census, we believe that ONS should work much
more closely with local authorities and harness their local expertise
and knowledge. If there are significant discrepancies between
the results of the Census and pre-existing data, ONS should engage
with local authorities to understand why this might be the case
before finalising and publishing results.
3.3 To achieve this, we are calling for
the 2011 Census to be planned for and carried out in partnership
with local authorities. Because of the importance of Census results
to service and policy development, we believe that local authorities
would welcome and support such a step.
4.1 The area of population and migration
is a specific example of how a partnership between ONS and local
government could be expected to add value to the 2011 Census process.
4.2 The decennial Census continues to provide
the only benchmark of the national population. In the years between,
annual population estimates for the UK and local authority areas
are calculated by ONS using a "components of change"
model which uses birth, deaths and migration data. Towards the
end of each inter-censul period, these estimates at the local
authority level become increasingly inaccurate. By 2001, the gap
between the ONS population estimate for Telford & Wrekin and
estimates independently produced by Prof Worrall at the University
of Wolverhampton was some 6,000 people.
4.3 Although this undercount was corrected
by the 2001 Census, there is already evidence that the current
series of ONS annual population estimates is undercounting Telford
& Wrekin's population. More recent research by Prof Worrall
has suggested that the ONS estimate for 2005 undercounted the
borough's population by some 2,000 people. Based on evidence from
the previous two decades, this undercount will only increase.
4.4 The reason why ONS annual population
estimates are unreliable is the quality of the migration data.
The failure of ONS to accurately estimate the number of migrants
into and out of the UK has been highlighted in a number of high
profile debates. As well as failing to accurately estimate
people entering and leaving the UK, ONS has failed to effectively
estimate the distribution of migrants into and out of local authorities.
This has resulted in authorities, such as Telford & Wrekin,
which have a high rate of population growth and relatively low
skill economies which are attractive to highly mobile migrants,
receiving annual population estimates from ONS which significantly
undercount their population.
4.5 Because of the importance of population
data to the Government's calculation of the revenue support grants
for local authorities, affected councils have received grants
which do not reflect the size of the population to which they
must provide services. We estimate that since becoming a unitary
authority in 1998, Telford & Wrekin Council has lost in
excess of £12.5 million in lost grant because of population
undercount by ONS. This shortfall in funding has put at risk
the ability of these local authorities to continue to deliver
the Government's sustainable growth agenda.
4.6 The 2011 Census provides ONS with a
unique opportunity to quantify and profile the level of migration
into the UK over the previous decade and importantly, provide
a benchmark for measuring and modelling migration for the forthcoming
decade. With this data, ONS should be better able to produce more
accurate annual population estimates.
4.7 ONS must specifically prepare for this
challenge and consider:
how to effectively engage migrants
in the Census process;
where to target Census survey resources
at new migrants.
4.8 Telford & Wrekin Council believes
that ONS can most effectively address these two issues by working
in partnership with local authorities. At present there is
no national dataset in the UK which has effectively tracked the
estimated 500,000 new migrants from EU accession states. Many
local authorities, however, through their own local research and
front-line community development and outreach workers are aware
of the neighbourhoods where these migrants are living. It is imperative
that ONS harnesses this local expertise to target survey resources
and ensure that the 2011 Census successfully counts both new and
highly mobile migrants.
5. CENSUS OUTPUTS
5.1 Part of the strength of the Census is
that it collects data about even the smallest communities across
the UK. We recognise that in planning the output and analysis
of this data ONS has to balance access to this detailed data against
data protection legislation and other appropriate safeguards around
individual privacy. However, a number of concerns were raised
over the way in which ONS managed the output of the 2001 Census.
5.2 The original ONS output plan for the
2001 Census was for a "one number" Census. In developing
this, ONS committed to ensuring that for any variable, all sub-authority
output tables (for example local authority wards) would sum to
separately published local authority level tables. With regret
this did not happen.
5.3 ONS developed their output strategy
for the 2001 Census very late in the process and relied on maintaining
anonymity of respondents by adjusting all table cells of value
3 or less to either a 0 or 3. As a result, data in sub-authority
level tables would not sum to the area total and different output
tables gave different results for apparently the same variable.
5.4 For larger groups of the population
this was less critical as the differences were proportionately
small, but for smaller communities the impact of this was much
greater and the value of the Census as a means to understand micro
local conditions was lost.
5.5 Telford & Wrekin Council believes
that the 2011 Census output strategy should be developed by ONS
as soon as possible and with the specific objective of ensuring
clarity. Anything other will significantly undermine the value
of this unique social survey.
5.6 In developing this strategy, we believe
that ONS should consider producing specific sets of output tables
for local authorities which are based on actual counts/estimates
and are not "distorted" in any way. Doing this would
require a very different relationship between ONS and local government,
however, it would ensure that the maximum value from the Census
would be achieved and that public resources could be targeted
at areas of greatest need.
6. ENSURING THE
THE 2011 CENSUS
6.1 Because of the uniquely detailed picture
the results will paint of local communities, the 2011 Census is
fundamental to policy development by both local and national government.
However, the 2011 Census will be a costly and challenging exercise
for ONS to undertake, particularly the need to maximise the response
rate from all sections of society and produce output of relevant
quality and consistency. If the 2011 Census does not meet these
objectives, the capacity of local and central government for the
following decade to develop effective social and economic policy
and target resources at areas of greatest need will be significantly
6.2 To ensure its success, we believe
ONS must avoid planning the 2011 Census as a top-down government
project and develop an approach which is based on partnership
with key stakeholders including local authorities. ONS will
require and should seek the local knowledge and expertise of local
authorities who are well placed to provide insight and understanding
of their local areas to ensure Census resources are appropriately
6.3 To achieve this, ONS should be planning
now to ensure that the appropriate local authority liaison networks
are in place and well embedded prior to the 2011 Census to ensure
both the fieldwork and output are of the highest possible standard.
Such a network would also provide the means by which ONS could
engage with local authorities if the 2011 Census results are out
of line with pre-existing data for local authority areas.