Memorandum from the Bank of England
EXECUTIVE SUMMARY
— There is considerable uncertainty
around current population estimates. This primarily relates to
estimates of international migration, since the last Census, which
are based on the International Passenger Survey (IPS). The numbers
of births and deaths are both well measured.
— The IPS only interviews a small number
of migrants at the main ports of entry into the UK. It was not
designed as a survey to measure migration and may not have been
able to fully keep pace with recent changes in migration patterns.
— Other administrative data sources
on National Insurance numbers issued and registrations with the
Worker Registration Scheme can be used as a cross-check against
the plausibility of official migration estimates. These alternative
sources suggest that there is a risk that the official estimates
of migration from the A8 countries in recent years could be too
low.
— The Office for National Statistics
(ONS) definition of population only includes migrants who come
to the UK for at least a year. Large numbers of individuals enter
the country to study or to work for a period of weeks or months.
These "visitors" will not be captured in the official
migration and population statistics, and they may contribute to
the level of demand and supply in the economy.
— Accurate estimates of the population
are important to the Bank in setting monetary policy since it
requires an accurate judgement regarding the amount of supply
available in the economy and the balance between this and the
demand for goods and services. This, in turn, may have important
implications for inflation.
— The Bank, along with other key users,
was consulted by the ONS when compiling the (inter departmental)
Migration Taskforce Report. The recommendations made in that Report
were very welcome, and the Bank continues to look forward to their
swift implementation.
— E-Borders data could be used in the
construction of migration and population statistics. For our purposes
it is not necessary to wait until full coverage of e-Borders becomes
available.
— The Bank would like to see the new
questions on the month and year of entry into the UK included
in the next Census.
MAIN SUBMISSION
Topic 1: Uses and definitions
What are the uses of population estimates, and
how far do current population estimates meet the needs associated
with those uses? What are the effects of inaccuracies or inadequacies
in such estimates?
1. Accurate estimates of the population
are important to the Bank in setting monetary policy since it
requires an accurate judgment regarding the amount of supply available
in the economy and, in turn the balance between this and the demand
for goods and services. This in turn, may have important implications
for inflation. Precise measurement of the total number of people
in the country is important, but accurate estimates of the age,
gender and regional distribution of the population are also important
since these more disaggregated estimates are used to weight responses
to the Labour Force Survey data and provide information on the
population share in work and the population share who are available
to work.
2. There is considerable uncertainty around
current estimates of population which creates difficulties in
judging the supply capacity of the economy. The issues relate
to estimates of net migration into the UK in the period since
the 2001 Census; the numbers of births and deaths are both well
measured. The uncertainty around official migration estimates
primarily relates to the International Passenger Survey (IPS)
in the construction of the data. The specific issues are discussed
in detail in paragraphs 11 and 12.
3. Uncertainty about the extent of migration,
and the characteristics of those migrants, makes it more difficult
for policy makers to judge the degree of inflationary pressure
in the economy. If, for example, immigrants raise aggregate supply
more than they raise aggregate demand, then one would expect inflationary
pressures to ease for a period of time. For further details see
the box entitled "The macroeconomic impact of migration"
in the November 2006 Inflation Report.[6]
4. It is possible that inaccuracies in population
estimates could lead to errors in monetary policy, and it is therefore
very important for the Bank that the population is measured precisely
and in a timely manner so that accurate information can feed into
policy decisions. For example, if the Monetary Policy Committee
(MPC) were to take a set of population estimates (or projections)
at face value which subsequently turned out to over or under-record
the number of people in the country, they may over or under-estimate
the supply capacity in the economy. Faced with a demand shock
the MPC may not change interest rates by as much—or sufficiently—as
was required to keep inflation at target.
5. But being able to identify the degree
of uncertainty in the most recent year for which there is official
population data is not a sufficient summary statistic on its own.
The operation of monetary policy requires us to take a view on
the likely future path of supply and demand, and so any difficulties
identifying the recent historical path of migration might also
feed into official projections. In other words, uncertainty in
the recent past potentially increases the uncertainty around our
growth and inflation projections as we look further forwards.
How appropriate is a definition of the population
based on the usually resident population in the context of the
needs of the users?
6. The definition of population used by
the ONS—consistent with the definition used by the UN—only
includes migrants who change their country of usual residence
for a period of at least a year. This may be appropriate for measuring
the long-run impact of international migration on the population
of the United Kingdom, but it may be too restrictive when thinking
about the economic impact of migration. Large numbers of individuals
enter the country to study or to work for a period of weeks or
months. These "visitors" will not be captured in the
official migration and population statistics, and they may contribute
to the level of demand and supply in the economy. It is quite
likely that the importance of these "visitors" will
have changed over time, with the advent of easier/cheaper travel
and, particularly, greater European integration. It is therefore
important to know how many of these short-term migrants there
are in the UK, how long they typically stay, and what their labour
market characteristics are.
7. To address this issue, the ONS have recently
published some preliminary estimates of the number of short-term
migrants in the UK.[7]
However, these estimates are also based on the IPS and they suffer
from the same problems as the estimates of long-term migration.
A very small sample size means that the data are highly uncertain.[8]
The Monetary Policy Committee also monitors a broader measure
of the number of people present in the UK at any one time that
is derived from flows of people into and out of the country by
air and sea.[9]
This measure will include short-term migrants, although it also
includes business travellers, tourists and other transient visitors.
The Worker Registration Scheme data are also used to help.
Topic 2: The role of the Census
How does the Census contribute to the creation
of population statistics? Is the current frequency of Censuses
appropriate and, if not, how frequent should Censuses be?
8. The Census makes an important contribution
to the creation of population statistics because it is the most
accurate measure of the UK population there is, and it provides
a benchmark against which the accuracy of the mid-year population
estimates produced since the previous Census can be evaluated.
The 2001 Census estimated that the UK population was about 1.2
million less than the mid-2001 projection.[10]
Much of this discrepancy is thought to have been a result of underestimation
of emigration. This prompted the ONS to launch a Quality Review
of migration statistics in an attempt to improve their estimates
of this source of population growth.[11]
In the light of the evidence suggesting significant increases
in net migration in the years after 2001, it therefore seems even
more critical that measurement issues that arose in the last Census
are appropriately dealt with in the next Census and that the Census
itself is sufficiently well resourced.
To what extent is there a trade-off between the
length of the Census form and its role in providing population
information? What questions should be included in the 2011 Census?
9. The Bank would like to see the new questions
on the month and year of entry into the UK included in the next
Census. These questions were included in the Census test that
took place in May 2007. The responses should help to improve the
quality of estimates of net migration since the previous Census,
around which there is currently a great deal of uncertainty. It
should be noted that a longer Census form may have the undesirable
consequence of creating a further delay in the release of the
population figures, which in the past have occurred with a significant
lag (of over two years).
Topic 3: Mid-year population estimates
How accurate and useful are the mid-year population
estimates for England and Wales, including information available
for local authorities and Strategic Health Authorities?
How appropriate is the methodology by which mid-year
population estimates are reached?
10. There is a great deal of uncertainty
surrounding the current mid-year population estimates. As noted
in paragraph 1, the uncertainty relates to the measurement of
net migration. Official estimates of international net migration
are primarily based on the International Passenger Survey (IPS),
a survey of individuals passing through the main UK air and sea
ports and the Channel Tunnel.[12]
The ONS supplement the IPS with administrative data on asylum
seekers and their dependents, and estimates of the migrant flow
between the United Kingdom and the Irish Republic; and the ONS
make adjustments to account for those whose intended length of
stay changes.
11. The IPS is a voluntary survey which
questions in excess of 250,000 travellers annually. Of those,
approximately 1% are migrant interviews. In 2005 (the last year
for which detailed data is available), the IPS statistics on migration
were based on interviews with 2,965 people who entered the United
Kingdom and 781 people who left. This is obviously a very small
sample and is one reason why there are large uncertainties surrounding
the official migration numbers. In addition, it is difficult for
the survey to keep pace with the dramatic change in the pattern
of arrivals seen in recent years, for example in the change in
movements between the UK and A8 countries.[13]
In 2005, only 94 citizens of the A8 countries were interviewed.
Approximately 90% of all migrant interviews took place at Heathrow,
with very few interviews taking place at other airports such as
Stansted and Luton. Other reasons why the IPS may mismeasure immigration
are that: the survey was not designed to measure net migration,
it was designed to capture tourism and business travel; participation
in the survey is voluntary and immigrants may be less likely to
respond (perhaps because of language barriers); and the survey
asks about intentions on arrival, not what people actually do.
12. There is a risk that current population
estimates could be under-recording the true population. Other
sources of administrative data suggest that net migration from
the A8 countries may be higher than is recorded in the latest
population estimates. The most recent official population data
suggest that between mid 2004 and mid 2006 there was a gross inflow
of 151,000 A8 citizens into the UK whose intended length of stay
was at least one year, and a net inflow of 131,000. Between May
2004 and June 2006, 433,000 A8 nationals registered for work under
the Worker Registration Scheme (WRS) and had their application
approved. This total will include temporary workers as well as
longer term migrants, but it will exclude those who are self employed,
students and dependents who are not working, all of whom would
be included in the official population data. Over the same period,
just over 400,000 National Insurance numbers were issued to people
from A8 countries. The reconciliation between the population data
and the other sources requires that either a very high proportion
of those registering for the WRS or applying for National Insurance
numbers are temporary migrants, or that the population data is
under-recording the true level of net migration from the A8 countries.
The Bank attaches some weight to both hypotheses.
13. The ONS have a work programme currently
in progress to address a number of the issues on the measurement
of migration highlighted above.[14]
Initial work includes increasing the sample of migrants at key
ports in the IPS and revising the assumptions on numbers of international
migrants whose actual length of stay differs from their stated
intentions. More substantial improvements are planned over the
longer term. The e-Borders programme will create a systematic
register of people entering and leaving the country which can
be used to improve the quality of migration statistics. Full coverage
of e-Borders is not expected for some time.
14. Population estimates at the local authority
level are likely to be less reliable than the aggregate population
estimates. They depend on assumptions about internal migration
within the UK and about which regions international migrants arriving
in the UK will settle in, both of which add additional uncertainty.
Internal migration is estimated using data on people changing
their GP as a proxy. This adds uncertainty to the estimates because
people are not obliged to re-register when they move and there
may to be differences in the patterns of registration among different
groups and in different periods. Estimates of international migration
at a regional level are likely to be even more uncertain than
at a national level. One of the key inputs is a question in the
IPS about where a migrant expects to live on arrival into the
UK. This question relates only to intentions as a migrant arrives
in the country, not where they actually settle, and again it adds
further uncertainty to the estimates.
What is the right process for making revisions
to mid-year population estimates so as to meet the needs of users?
15. Revisions should be made to population
estimates when the accuracy of the statistics can be improved.
Where methodological improvements are introduced, revisions should
be made to the back data as far back as necessary to ensure that
consistent comparisons of population can be made over time.
Topic 4: The role of survey and administrative
data
What role can and should be played by survey data
in the compilation of population statistics?
16. Sample survey data is required in the
compilation of population statistics where comprehensive survey
data such as the Census or administrative data sources are not
available. Sample survey data should be benchmarked against Census
type data at regular intervals to evaluate the accuracy of those
survey results.
What role can and should be played by data from
General Practitioners and other health service data in the compilation
of population statistics?
17. Data from General Practitioners is already
used in producing population estimates at a local authority level
where estimates of internal migration are required. This appears
to be the best source of data available for measuring internal
migration in between Census years. GP records on the numbers registering
who previously lived abroad could be used as a cross-check on
the international migration data, and particularly on the regional
distribution of international migrants. However, a major limitation
of doing this is that the country and region of origin are not
collected in the GP records and only those wishing to access NHS
services through a GP are likely to register.
What other data sources can and should be used
in the compilation of population statistics?
18. There are a range of other data sources
available on international migration which may be of use in the
construction of population data. National Insurance numbers issued
and the Workers Registration Scheme data for A8 nationals are
likely to be the most useful sources. Until the arrival of e-Borders
there is no obvious replacement for the IPS as the main survey
used to measure migration, but given the limitations of this survey
already discussed other data sources could be used as a cross-check
on the plausibility of IPS based migration data. The main problem
with the alternative data sources is that they use different definitions
of a migrant to that used in the mid-year population estimates
and many can only be used to measure gross inflows rather than
net migration. The ONS have already carried out a review of the
potential to use these administrative data sources in population
estimates.[15]
A publication to bring together and report in a coherent manner
all the statistics collected across Government on migration and
migrants is also planned.[16]
The ONS also plan a migration module, as an ad hoc addition to
the LFS in 2008.
19. Going forward, e-Borders data could
be used in the construction of migration and population statistics.
For our purposes, it is not necessary to wait until full coverage
of e-Borders is achieved. The data could be used as soon as a
sample that is large enough to be representative is available.
Topic 5: Cooperation with stakeholders
How effectively has the Office for National Statistics
cooperated with stakeholders with an interest in and information
relating to population statistics and how can cooperation be improved
by the Statistics Board and by its Executive Office?
20. The ONS cooperates well with the Bank
in relation to population statistics. We have quarterly liaison
meetings at which issues relating to population and migration
statistics can be and are regularly discussed. The Bank, along
with other key users, was consulted by the ONS when compiling
the Migration Taskforce recommendations. The Bank welcomes the
recommendations of the Taskforce, and continues to look forward
to their swift implementation. The Bank has also been consulted
on its need for migration information from the 2011 Census.
January 2008
6 See http://www.bankofengland.co.uk/publications/inflationreport/ir06nov.pdf Back
7
See http://www.statistics.gov.uk/about/data/methodology/specific/population/future/imps/updates/
downloads/STM_Research_Report.pdf Back
8
The estimates for the year to mid-2005 are based on only 120 interviews
with migrants flowing into the country and 38 interviews with
migrants leaving. Back
9
For example, see p27 of the February 2007 Inflation Report. Back
10
See the box entitled "The 2001 Census" in the November
2002 Inflation Report for further details. Back
11
For full details see http://www.statistics.gov.uk/about/data/methodology/specific/population/future/imps/updates/default.asp Back
12
For further details see box entitled, "International migration
data", in the August 2005 Inflation Report, http://www.bankofengland.co.uk/publications/inflationreport/ir05aug.pdf. Back
13
On 1 May 2004, 10 countries joined the EU. They are the A8 countries
(Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland,
Slovakia, Slovenia) whose citizens gained the right to live and
work in the UK, plus Cyprus and Malta whose citizens already had
the right to live and work in the UK. Back
14
For full details see http://www.statistics.gov.uk/about/data/methodology/specific/population/future/
imps/updates/default.asp Back
15
See http://www.statistics.gov.uk/about/data/methodology/specific/population/future/imps/updates/downloads/
admin.pdf Back
16
This was a recommendation of the Migration Taskforce Report. Back
| 
