Migration Statistics - Public Administration Committee Contents

3  Local area migration estimates

37. The ONS produces estimates of international migration to and from local authority areas. These are based principally on the same IPS data that are used to estimate migration for the UK as a whole, which are adjusted using administrative data.[31] Users of migration statistics identified local area migration statistics as being particularly weak. This is despite the ONS taking steps to improve local area migration statistics as part of its Migration Statistics Improvement Programme. The Royal Statistical Society told us:

    There is a continued problem with the quality and quantity of migration data available at a local level. Improvements have been made through the Migration Statistics Improvement Programme by allocating international migrants to local authority areas using administrative data, but the local-level estimates of migration are not robust, particularly for areas with high population turnover. For some local authority areas the Census has shown that the ONS population estimates have misrepresented the level of population growth, a problem caused by inaccurate internal and international migration estimates.[32]

38. The population of England and Wales recorded at the 2011 Census was 476,000 larger than the ONS expected, based on its methodology for estimating the population between Census years. The ONS attributed 267,000 of this difference to inaccuracies in the measurement of net international migration over the decade to 2011, of which around 175,000 was attributed to systematic underestimation of net migration by the IPS.[33]

39. The ONS told us the total discrepancy between the Census and the estimated population in March 2011 was just 0.8% of the population of England and Wales. They suggested this was "not bad",[34] although in absolute terms it represents a population equal to the city of Liverpool. In some local authorities the discrepancy between the Census and the expected population was much greater as a proportion of the population than for England and Wales as a whole. In around 11% of local authorities the difference was greater than 5% of the population, and in 9% of local authorities it was greater than 10,000 people. In the London Borough of Brent, the population was around 35,000 larger than expected, which was more than 10% of the expected population, while in some local authorities the population was smaller than expected.[35]

40. The Oxford Migration Observatory drew attention to the fact that there is no source of data on migration in local areas that is both complete and accurate other than the decennial Census, which rapidly becomes out of date and may not continue in its current form. The ONS's "Beyond 2011" programme is examining options for the future of the Census. PASC will conduct its own study into the Census as part of its wider programme of work on statistics and their use in Government. The Observatory also drew attention to the difficulties caused by the lack of robust data on migration at the local level:

    The significant uncertainty about the number of migrants in local areas creates significant difficulties for the planning and efficient delivery of public services and a whole range of other public policies.[36]

41. Westminster City Council was particularly critical of migration estimates for local authorities, especially in their effect on local area population estimates. It said the migration estimates were unreliable and complained that frequent revisions to the estimated local authority population made it difficult to plan services. It argued that the current methodology for estimating migration was not robust enough to support accurate local level estimates, and it was concerned that the unreliability of the estimates "will cause severe detrimental impacts in the future grant settlements." It said recent changes to the methodology for estimating migration in local areas had "not improved the measurement of migration for places like Westminster", and it concluded that "the measurement of migration from the perspective of a LA [local authority] user and as reliable information on our residents is failing".[37]

42. The ONS told us the factors leading to the underestimation of net migration in the decade to 2011 had largely been addressed through improvements to the local area migration estimates introduced in 2009 as part of the Migration Statistics Improvement Programme. It argued that, had these improvements been introduced at the start of the last decade, "the population estimates would have been much closer to what the Census said".[38] But this claim is untested. The extent of any systematic error in the new methodology for estimating international migration in local areas is unknown.

43. The basis for estimates of international migration in local areas is IPS data on migrants' origins and intended destinations within the UK. The weakness of these estimates is mainly due to the size of the IPS sample and its inability to provide detailed information on the characteristics and behaviour of migrants.[39] In addition, many immigrants do not know their ultimate destination upon arrival in the UK. All of the problems with migration estimates at a national level are magnified at the local level.

44. In its recent review of the robustness of the International Passenger Survey, UKSA concluded that:

    The IPS sample size is too small to enable the production of reliable international migration estimates at a local authority level, and cannot realistically be made sufficiently large to achieve robust local estimates.[40]

In oral evidence, the ONS acknowledged the weakness of the IPS as a source of data on migration in local areas.[41]

45. The International Passenger Survey does not provide accurate estimates of international migration in local areas. The Census provides the most accurate data on the number and characteristics of migrants at the local level, but it is too infrequent to act as a routine source of data. The future of the Census is also uncertain. As the only reliable source of data on migrant populations in local areas, the potential loss of the Census is a concern. Accurate estimates of migration in local authorities must be available independent of the Census. The ONS should develop new sources of data on international migration that are robust enough to provide accurate estimates of annual migration flows to and from local authority areas, even if the Census continues.

31   ONS, Improved Methodology for Estimating Immigration to Local Authorities in England and Wales, November 2011 Back

32   Ev w18 Back

33   ONS, Explaining the Difference between the 2011 Census Estimates and the Rolled-Forward Population Estimates, September 2012, pp 2-8 Back

34   Q 60 [Mr Humberstone, ONS] Back

35   ONS, Explaining the Difference between the 2011 Census Estimates and the Rolled-Forward Population Estimates, September 2012, p 12 Back

36   Ev w10 Back

37   Ev w22 Back

38   Q 86 [Mr Humberstone, ONS] Back

39   See Chapter 2 above. Back

40   UKSA, Monitoring Brief reviewing the robustness of the International Passenger Survey, June 2013, p 5 Back

41   Q 59, Q 61 Back

previous page contents next page

© Parliamentary copyright 2013
Prepared 28 July 2013