Migration Statistics - Public Administration Committee Contents

2  International migration estimates


5. International migration estimates produced by the ONS attempt to reflect how much the UK population has changed as a result of migration. They estimate how many migrants have been arriving and leaving, where they are migrating to and from, and for what reasons. They aim to measure all long-term international migration to and from the UK, which is migration by people who change their country of usual residence for a period of at least a year. ONS migration estimates are the only statistics that attempt to routinely measure all migration to and from the UK within a given period using a consistent definition. As migration is a component of population change, ONS migration estimates are used in estimating population change in years between the decennial Censuses.

6. ONS migration estimates are based principally on the International Passenger Survey (IPS), which is a survey of passengers arriving at and departing from UK air and sea ports. The IPS was not primarily designed for the purpose of estimating international migration, but to provide economic data on travel and tourism. As most people travelling to and from the UK are not long-term international migrants, the IPS must approach around 800,000 passengers each year in order to achieve a sample of around 5,000 migrants.[3]

7. The ONS first produces estimates of international migration using just the IPS sample. These survey-based estimates are then adjusted to reflect types of long-term migration that are not properly captured by the survey, such as asylum seekers, migration through Northern Ireland, and "switchers"—people who change their country of usual residence for a period longer or shorter than they originally anticipate, thereby falling into and out of the definition of a long-term migrant. The adjusted estimates are the long-term international migration (LTIM) estimates, which provide the headline estimates of immigration, emigration and net migration in the UK.[4]

8. The ONS has produced long-term international migration estimates for each calendar year since 1991, for years to June since 2003, and for years to March and September since 2010. Figure 1 shows LTIM estimates of immigration, emigration and net migration in years to December and June since 1991, for those years where data are available.

Figure 1: Long-term international migration, Years to Dec & Jun, 1991-2012

Note: Years to June only available since the year ending June 2003.

Source: ONS, Migration Statistics Quarterly Report, May 2013

9. Net migration is the difference between immigration and emigration: the number of people moving to live in a particular country minus the number of people moving out of that country to live elsewhere. It is important to recognise that net migration does not indicate the full extent of population change. If immigration and emigration are roughly equal, net migration will be low irrespective of how many people are migrating in and out of the country. Low net migration can be consistent with high levels of immigration and emigration, and net migration may change as a result of changes in either immigration or emigration.

10. Net inward migration to the UK increased from an annual average of around 37,000 in the period 1991 to 1995 to an annual average of around 209,000 in the period 2006 to 2010. So while the Government's net migration target—to reduce net migration from the hundreds of thousands to the tens of thousands—suggests a ten-fold reduction in net inward migration, in practice net inward migration only needs to be roughly halved in order for the Government to achieve its aim.

11. Because the international migration estimates are based on a survey sample, rather than a count, they are subject to a margin of error. In practice this error can be quite large. For net migration (as measured by the unadjusted IPS estimate), the range is currently around plus or minus 35,000, which means there is a 95% chance that the true level of net migration falls within a range of around 70,000. This is called "the confidence interval" and is calculated by ONS statisticians. Confidence intervals cannot be calculated for the final adjusted LTIM estimates because the uncertainty associated with some of the adjustments cannot be easily quantified, but the confidence intervals surrounding the IPS estimates give some indication of the statistical uncertainty in ONS migration statistics. This means that if the estimate of net migration is, for example, 200,000, the confidence interval suggests there is a 95% chance the true value falls between 165,000 and 235,000, with a 5% chance the true value falls outside this range. In practice, the uncertainty is even greater than this, as the confidence interval only represents the potential sampling error, and takes no account of other possible sources of error, such as the possibility of respondents lying or systematic biases in the willingness of particular groups to participate in the survey.

12. Figure 2 shows immigration and emigration in the year to June 2012 broken down by nationality. Just over half of immigration to the UK is by non-EU nationals (55%), around a third is by nationals of EU countries other than the UK (30%), and less than a sixth is by British nationals (15%). This means that just over half of the immigration flow comprises people who need a visa to come to the UK. Conversely, most emigration from the UK is by British nationals (44%), around a third is by non-EU nationals (32%), and around a quarter is by nationals of other EU countries (24%).

13. Figure 3 shows net migration by nationality in years to June from 2003 to 2012. Net immigration is highest among non-EU nationals, while there has been net emigration of British nationals in every year since LTIM estimates have been produced. Because more British nationals are leaving the UK than entering it, the overall level of net migration is lower than would otherwise be the case. An increase in net emigration of British nationals since 2010 has contributed to the fall in net migration during this period.

Figure 2: Immigration and emigration by nationality, Year to June 2012

Source: ONS, Migration Statistics Quarterly Report, May 2013

Figure 3: Net migration by nationality, Years to June, 2003-2012

Source: ONS, Migration Statistics Quarterly Report, May 2013

Quality of international migration estimates

14. In 2009 the UK Statistics Authority (UKSA) carried out an assessment of ONS migration statistics. It said that "Both users and ONS' statisticians generally agree that migration statistics are not fit for all of the purposes for which they are currently used and require further improvement".[5]

15. Between April 2008 and March 2012 the ONS carried out a comprehensive programme for improving the quality of its migration statistics. The Migration Statistics Improvement Programme (MSIP) led to some improvements in the way migration statistics are produced and reported.[6] The ONS told us the migration statistics were "fit for purpose".[7]

16. In written evidence, users of ONS migration statistics welcomed these improvements, but many remained critical of the overall quality of the international migration estimates. The Royal Statistical Society told us:

    Despite these recent improvements migration statistics are still not fully adequate for the task of producing robust population estimates or understanding patterns of migration.[8]

The British Society of Population Studies wrote:

    The statistics on migration to and from the UK and its constituent parts are inadequate not only for social scientific inquiry but also for monitoring the effectiveness of measures designed to implement government policy.[9]

The Royal Geographical Society said "the international migration data are not fit for purpose", but concluded that "the ONS (and its sister agencies) are doing a good job with poor data".[10]

17. The most common criticisms of the migration estimates were the degree of uncertainty surrounding the statistics and the lack of detailed information available on the characteristics and behaviour of migrants entering and leaving the UK.[11] Both of these problems were attributed to the IPS and its small migrant sample. The British Society of Population Studies told us:

    [I]t can confidently be stated that the key problem with the quality of these migration statistics is the reliance on the IPS for the main element of the total numbers of immigrants and emigrants. Therefore the most obvious way of improving their quality is by reducing the degree of uncertainty surrounding the IPS-based estimates, which requires greatly increasing the number of migrants interviewed from its current level of around 12 a day.[12]

18. Many of those who submitted written evidence hoped that data from the Government's e-Borders programme—which records basic travel document information on the identity of passengers travelling through UK ports—could eventually be used to improve the quality of migration estimates.[13]

19. The ONS explored the potential use of e-Borders data as part of the Migration Statistics Improvement Programme. The biggest contribution they expect e-Borders data to make is in improving the accuracy of the headline measures of immigration, emigration and net migration at the national level. This improvement is expected around 2018, three full years after the e-Borders scheme achieves 95% coverage of passenger movements in and out of the UK.[14] This is because e-Borders data must be collected for a full year before and after the year for which migration statistics are produced in order to determine whether a person entering or leaving the UK meets the definition of a long-term international migrant.

20. However, while e-Borders data may lead to more accurate headline estimates of immigration, emigration and net migration, it cannot replace the IPS in providing information on the characteristics of migrants, as it does not record all of the characteristics of migrants that the IPS currently records, such as their usual occupation before migration or their main reason for migrating. It can also play only a limited role in helping to improve local area migration estimates as it does not record the origin or intended destination of people migrating to and from the UK. We look at this issue further in Chapter 3 below.

21. Users and producers of migration statistics both said that data on emigration was even weaker than data on immigration. Several respondents to our call for evidence suggested that emigration statistics could be improved by using data on immigration in other countries.[15] The Royal Statistical Society said:

    Estimates of emigration from the UK are known to be hardest to produce. The potential for use of other countries' immigration data should be considered to validate the UK estimates. Furthermore, the Office for National Statistics could proactively encourage cooperation between member states of international organisations such as the European Union, OECD and UN to work together on this issue.[16]

22. In a recent review of the robustness of the International Passenger Survey, UKSA also recommended using data on international migration collected in other countries to help understand migration to and from the UK.[17]

23. We welcome work the ONS has done to improve the quality of migration statistics. The ONS has done its best to produce informative migration statistics using the International Passenger Survey. However, the International Passenger Survey is inadequate for measuring, managing and understanding the levels of migration that are now typical in the UK. The Government must plan to end reliance on the International Passenger Survey as the primary method of estimating migration: it is not fit for the purposes to which it is put.

24. e-Borders data has the potential to provide better headline estimates of immigration, emigration and net migration from 2018. The ONS and Home Office should move as quickly as possible to measuring immigration, emigration and net migration using e-Borders data.

25. Migration is an international phenomenon. Data held by other countries on migration to and from the UK could help improve the depth and quality of UK migration statistics. The ONS should co-operate further with foreign national and international statistics agencies to improve the quality of UK migration statistics.

Measuring progress against the net migration target

26. In written evidence, many users of migration statistics expressed concern that the degree of uncertainty in the migration estimates made them unsuitable for measuring progress against the Government's net migration target. Migration Watch told us:

    The uncertainty around the net migration figure should be reduced. The net migration figure is central to the Government's policy on immigration and their success in this area will be largely judged on this figure. It is therefore far from ideal that the true net migration figure could deviate so substantially from the calculated estimate.[18]

The Oxford Migration Observatory wrote:

    The available migration estimates are problematic as a means to define and precisely measure progress toward a numerical limit on migration [...] For the Government to be judged on its achievement in delivering this target, accurate measurement is important. But to know whether this target has been reached requires clear data—of the sort that the IPS does not currently produce because of the uncertainty surrounding the estimates. As a consequence, the Government could miss the "tens of thousands" target by many tens of thousands and still appear to have hit it—conversely the Government could hit, or even exceed its target and still appear to have missed it by tens of thousands.[19]

Submitting evidence in his capacity as Chair of the Universities UK working group on student visa issues, Professor Edward Acton said that such uncertainty could lead to immigration policy that was either too tight or too loose.[20]

27. The only respondent to the Committee's call for evidence that felt the current migration estimates were adequate for this purpose was the Migration Advisory Committee (MAC), which advises the Home Office on migration policy.[21] However, it is worth noting that the MAC's recommendations to the Home Office have been influenced by the potential inaccuracy of the migration estimates. When the MAC was asked to suggest visa limits to help achieve the net migration target in 2010, it aimed at reducing net migration to 50,000 rather than 100,000 specifically in order to overcome uncertainty in the net migration estimate.[22]

28. We wrote to Sir Andrew Dilnot, Chair of UKSA, to ask whether UKSA considered the LTIM estimate of net migration to be suitable for measuring progress against the Government's net migration target. In his response, Sir Andrew wrote:

    With careful analysis of all the available data we can be fairly sure of the broad level of net inward migration over a period. It may, however, be necessary to wait quite a long time to get a clear picture.[23]

As illustrated in Figure 1 above, the estimated level of net migration can change by tens of thousands in either direction from one calendar year to the next, so medium and long-term trends in net migration only become apparent after several years. The true level of net migration at a given point in time is highly uncertain.

29. In the longer term, migration estimates based on the International Passenger Survey are too uncertain for accurate measurement of progress against the Government's net migration target. We are struck by the advice of the Migration Advisory Committee to the Government that it should aim for net migration of only 50,000 as the only means of being certain that net migration is in fact below 100,000. The Government should not base its target level of net migration on such an uncertain statistic: doing so could lead to inappropriate immigration policy.

Understanding who is migrating to and from the UK

30. The statistical uncertainty associated with migration estimates increases (relative to the size of the flows) when the IPS sample is broken down to identify particular sub-groups of migrants, such as those of a particular nationality, or those with a particular reason for migration. So, for example, when the estimate of net inward migration by nationals of the A8 Eastern European countries that joined the EU in May 2004 is 28,000, the 95% confidence interval surrounding the estimate is around plus or minus 14,000.[24] This means there is a 95% chance that the true number of migrants from those countries falls between 14,000 and 42,000.

31. This statistical uncertainty limits the extent to which the IPS sample can be broken down into different sub-groups. Therefore, breakdowns of immigration and emigration by nationality, country of birth, and country of last or next residence are only available for groups of countries (such as the EU15, the A8, and the Old and New Commonwealth) rather than for individual countries. In its evidence to this Committee, the Royal Statistical Society said:

    Changing patterns of international migration over recent years—and probably into the future—means that there is an increasingly wide matrix of countries of interest, for example the BRIC countries (Brazil, Russia, India and China). Yet the ONS continues with the same breakdown as in recent decades (Old and New Commonwealth, Other Foreign, EU etc.).[25]

32. Furthermore, as UKSA indicated in its written evidence, certain characteristics of migrants are not recorded by the IPS at all; such as ethnicity, religion, language, and educational qualifications.[26] The Royal Statistical Society noted that the absence of data on ethnicity in particular "makes it difficult to estimate the impact of international migration on the ethnic composition of the population, needed for planning of various services such as education and health".[27]

33. The ONS told us the IPS was not an appropriate source of data for detailed information on the characteristics of migrants.[28] It did not recommend trying to improve the level of detail available in migration estimates by increasing the size of the IPS, because even a very large increase in the number of people surveyed would not provide the level of detail users of migration statistics need. For this reason, the ONS argued that increasing the IPS would be "poor value for money".[29] Instead, it recommended that migration statistics could be improved more cost-effectively by developing new sources of data on migration, such as by fully integrating the IPS with e-Borders and by conducting specific routine surveys of migrants living in the UK.[30]

34. Migration estimates based on the International Passenger Survey do not provide sufficient detail on the characteristics of people migrating to and from the UK to judge properly the social and economic consequences of migration and the effects of immigration policy. These data are indispensible for anticipating demand for public services such as schools and the NHS. Migration statistics should provide detailed information on the characteristics of people migrating to and from the UK within particular periods, including information that is relevant to evaluating the impact of immigration policy and necessary for planning services. The ONS should broaden the information it gathers on the characteristics of migrants to include level of educational qualification, labour market skills, ethnic group, and languages spoken.

35. e-Borders data could potentially provide detailed information on the characteristics of migrants subject to visa control. However, e-Borders data alone will not provide detailed information on the characteristics of those migrants not subject to visa control, or any information on the geographical origin and destination of migrants within the UK. If the International Passenger Survey is not an adequate source for this information, and no other sources are available, new sources of migration statistics are needed, even though they may come at some cost.

36. The ONS should develop new sources of data that can provide accurate statistics on the numbers and characteristics of people migrating to and from the UK, and on their areas of residence within the UK. The ONS should link International Passenger Survey responses to e-Borders data as soon as possible. However, the need for further data may also require the creation of a new routine migrant survey covering the whole of the UK.

3   Q 59 Back

4   ONS, Long-Term International Migration Estimates - Methodology Document - 1991 onwards, May 2013 Back

5   UK Statistics Authority, Assessment Report: Migration Statistics, July 2009, p 7 Back

6   ONS, Migration Statistics Improvement Programme Final Report, March 2012 Back

7   Q 59 Back

8   Ev w18 Back

9   Ev w2 Back

10   Ev w1 Back

11   Ev w2, Ev w6, Ev w8, Ev w13, Ev w18 , Ev w27  Back

12   Ev w2 Back

13   Ev w1, Ev w2, Ev w8, Ev w10, Ev w13, Ev w18 Back

14   ONS, Delivering statistical benefits from e-Borders, p 7 Back

15   Ev w2, Ev w18, Ev w31 Back

16   Ev w18 Back

17   UKSA, Monitoring Brief reviewing the robustness of the International Passenger Survey, June 2013, p 3 Back

18   Ev w8 Back

19   Ev w10 Back

20   Ev w6 Back

21   Ev w8 Back

22   Migration Advisory Committee, Limits for Tier 1 and Tier 2 for 2011/12 and supporting policies, November 2010, paragraphs 45-48, p 12 Back

23   Ev w36 Back

24   The A8 countries are the eight Eastern European countries that acceded to the European Union in May 2004, comprising Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, and Slovenia. Back

25   Ev w11 Back

26   Ev w31 Back

27   Ev w18 Back

28   Q 59 Back

29   Q 61 Back

30   Q 63 Back

previous page contents next page

© Parliamentary copyright 2013
Prepared 28 July 2013