Migration statistics

Written evidence submitted by the British Society for Population Studies (BSPS)

The British Society for Population Studies (BSPS) comprises persons with a scientific interest in the study of human populations. Its main objectives are to further the scientific study of biological, economic, historical, medical, social and other disciplines connected with human populations and to contribute to the public awareness of them.

Summary

· 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.

· This remains the case despite the significant improvements in data coverage and accuracy achieved by the UK’s three national statistical agencies over the past decade following the revelations of the 2001 Census.

· The main reason for this unsatisfactory situation is that the primary source used in the monitoring of the UK’s total migration flows between the decennial checks provided by the Census does not have this as its primary purpose.

· In the absence of an official system of registering the addresses of all persons usually resident in the UK and the changes in these, the best way of improving the quality of the UK’s migration statistics is to implement e-Borders in such a way as to provide a full count of international arrivals and departures and, alongside this, to use the enhanced International Passenger Survey, Annual Population Survey and Population Census to estimate their personal characteristics and their UK destinations and origins respectively.

· The statistics on departures (emigrants) are now recognized to be of lower quality than those on arrivals (immigrants), yet are just as important in the calculation of the net migration figures that have featured recently in government policy statements and are just as important in their impact on the UK’s population size and composition.

· Additional steps needed for the more accurate measurement of departures include: systematic collection of data from non-UK statistical agencies on the numbers of people arriving in their countries from the UK; fuller examination of the UK’s three Longitudinal Studies to identify people that cannot be traced from one Census to the next and have not been registered as deaths; and better documentation of UK nationals living abroad.

Overview

1. The BSPS members who supplied evidence for this submission to PASC appreciate the substantial progress made in the UK’s international migration statistics over the last decade. The Migration Statistics Improvement Programme (MSIP) has led to refinements in the International Passenger Survey (IPS) and use of administrative data, greater availability of and accessibility to data, and more sensible publication dates, as well as providing valuable new information on plausibility ranges and migration timelines.

2. Nevertheless, there remain great concerns about the quality of the UK’s international migration statistics and the extent to which they are fit for the purpose of helping to understand and forecast this key driver of UK population change and to measure its impacts, especially at the local-area scale. Therefore they still cannot be deemed to meet their users’ needs adequately. This situation is unlikely to change without further investment in the IPS system or the adoption of e-Borders or an alternative system of better measuring the total numbers of people moving into and out of the UK.

Answers to PASC’s specific questions

1. Do the published migration statistics – at the national, regional and local levels – meet the full range of their users’ needs, namely:

a. Are they easily discoverable and accessible to all users?

3. The recently revamped ONS website is a distinct improvement on the previous version. Even so, it is not easy to negotiate for an experienced professional user, let alone for an interested layperson. Moreover, Google Search can sometimes get more directly to the data that one is searching for than using the ONS website’s own search facility.

b. Are they easy to use and understand?

4. The international migration statistics published by ONS constitute a large suite of different data sets which are easy for the experienced statistician to use and understand but have the potential to wrongfoot the unwary.

5. Most straightforward are the total long-term international migration estimates which are published as part of the ‘components of change’ statistics released with the annual population estimates down to the local authority scale. Even so, it is not easy to switch between data, methodology and interpretation. It is recommended that an interface be developed along the lines of the Neighbourhood Statistics where data can be accessed in raw format and the metadata is placed alongside.

6. The less experienced user needs to be aware that those published migration statistics which are based on data derived solely from the IPS do not provide the full picture of this long-term migration. There is also the potential for confusion arising from the ONS’s attempts – prompted in large part by the rise of labour migration from the EU’s new Accession States from 2004 – to measure short-term international migration, defined as people moving into and out of the UK to live for between 3 and 12 months. However, wisely ONS has been careful to keep these separate from its main statistics on international migration and out of the annual population estimates, giving central government the ability to allow for these separately in any relevant funding distributions.

c. Do they provide an appropriate level of detail?

7. Users vary considerably in their needs for detailed information about international migrants. Some are most concerned about the headline figure of the UK’s total net migration in a particular year, which is readily available from the published data.

8. Most users are keen to distinguish immigrants and emigrants separately by nationality (especially British versus non-British citizens), area of origin and destination (not just broad world regions but also individual countries especially for the main suppliers and receivers), the main purpose of moving (especially work, study, and family reasons) and personal characteristics such as sex, age, labour market skills, race/ethnicity, wealth and health. ONS publishes the data which is provided by the IPS, but this covers only some of these details. Other sources such as the Annual Population Survey (APS) and the Population Census can be used to supplement this picture, but this task is left to users and can cover only immigrants, and not emigrants, because these surveys are of current residents.

9. In addition, a substantial body of users – notably in local administration and planning but also in social science research – requires these types of details for small areas rather than just for the UK as a whole. For these, despite the advances made in recent years through the MSIP, the published data are unsatisfactory in two respects. Firstly, the lowest level at which data derived from the IPS and APS is normally released is the local authority area, which can be as large as one million people (e.g. Birmingham) and includes a number of unitary counties covering very extensive areas (e.g. Northumberland). Secondly, the quality of the data is much lower for this more detailed geography than for national level.

10. The only data on international migration that is published by ONS for small areas (i.e. below local authority level) is that derived from the Population Census. The latter provides high-quality estimates of the number of local residents who had been living outside the UK one year previously and information on their personal and household characteristics. Moreover, through its question on country of birth, the Census also gives the total stock of immigrants in each area and – in the 2011 Census for the first time – allows this to be broken down by year of arrival in the UK. On the downside, the Census cannot directly measure emigration (though see para 24) and provides a snapshot only once every 10 years.

d. Are they effectively summarised?

11. ONS does a good job in collating its statistics on international migration and providing commentary on the main features that they reveal. Its annual report International Migration has traditionally contained a mine of information primarily drawn from the IPS. Particularly helpful currently is the Migration Statistics Quarterly Report which also contains summaries of the data obtained from the Home Office, as also is the Local Area Migration Statistics data file which includes data on immigration derived from other sources such as the Workers Registration Scheme and National Insurance. The ONS website also contains relevant items, though these are usually brief: ONS should not abandon the practice of preparing longer articles along the lines of the annual reports that used to be published in Population Trends.

2. How well have producers of migration statistics engaged with users? How responsive have they been to feedback from users of statistics?

12. Over the years ONS has built up an impressive record of engagement with users of migration and related statistics. In particular, it involved relevant experts in the MSIP and in planning the 2011 Census, which provided much more intelligence on immigrants than any previous one. In addition, ONS regularly seeks the views of users through its formal consultations and its presentations at roadshows and conferences, including BSPS meetings.

13. ONS is also considered to be highly responsive to feedback, both in taking on board the suggestions received from users and in giving clear reasons in cases where they have not felt able to do so. The latter situation has tended to occur only when there have been significant technical and/or financial barriers to meeting user demands.

3. Do the migration statistics which are published enable members of the public to gain a better understanding of the issues? Are the right migration statistics being collected?

14. If the term ‘issues’ refers to the size of the UK’s overall net migration balance (as in question 5), then the right type of migration statistics is being collected and published for the purposes of allowing the public to see how this number is changing over time. Migration statistics alone, however, are not sufficient to enable a better understanding of whether a higher or lower number, or some other change in the patterning of this migration, would be beneficial to the UK as a whole or to the individual person or the section of the population to which they belong. Such an understanding can come only from the results of research on the causes and consequences of all the various population movements that are included under the headings of immigration and emigration, these including skilled and unskilled labour migration, student migration, retirement migration, refugees and asylum seekers, and family reunification. The responsibility for providing this sort of intelligence to the public lies beyond the remit of the statistical agencies. The main question for the latter is whether the published statistics are correct, i.e. provide an accurate record of what is actually happening (see below).

4. Is the degree of uncertainty surrounding estimates of migration properly reported and widely understood? Is the degree of uncertainty surrounding estimates of migration acceptable or should it be reduced? If so, how could it be reduced?

15. While the ONS website provides clear guidance on the degree of uncertainty that is attached to its estimates of international migration, this is not nearly so evident in the summary publications that are most accessible to the media and the public at large. If it were, then surely there would have been greater pressure placed by the electorate on government to improve the quality of the statistics on international migration. This is because the average person would find the scale of uncertainty difficult to comprehend.

16. As an illustration, the latest edition of ONS’s Migration Statistics Quarterly Report (November 2012) contains one paragraph on uncertainty, specifically to say that confidence limits have been introduced to accompany the migration estimates based on the IPS. Using the web link given there and then being directed through several further web pages, it is found that the central estimate of net international migration for 2011 is put at 199,600. The 95% confidence limit, given as 35,400, indicates that it is very likely that the true figure lies between 164,200 and 235,000, with a 1 in 20 chance that it lies outside this range. This degree of uncertainty is mainly because the central estimates of arrivals and departures – 531,300 and 331,600 in 2011 – are based on very small sample sizes: interviews with just 2,620 and 1,824 people respectively in 2011.

17. Ultimately, whether or not this degree of uncertainty is acceptable has to be a political decision, informed by an appreciation of the importance of any government target for managing international migration weighed against the costs of introducing measures to reduce the uncertainty. It is, however, worth saying that there would be much cause for concern if the statistics on the other two basic determinants of the UK’s changing population size – births and deaths – were subject to a similar level of uncertainty.

18. How it could be reduced is dependent on improving the quality of migration statistics (see question 6).

5. Are the migration statistics adequate for measuring the Government’s progress against its net migration target?

19. No, as exemplified in the answer to question 4. But then the Government’s net migration target is poorly conceived: indeed, it provides a very good illustration of the prevailing lack of understanding of migration. There seems to be a general failure to recognise that the net figure is the balance between the two much larger figures of immigration and emigration. While most public concern seems to be focused on the number of immigrants, the Government’s net target could in theory be achieved without any change in the number of people moving to live in the UK but instead by a 30 per cent increase in the number of people moving abroad.

6. What more could be done to improve the quality of migration statistics? Should data from other sources, such as e-Borders, be incorporated?

20. In the last few years changes have been made in the methods of measuring the numbers of people entering and leaving the UK for intended stays of at least one year and also in the methods of allocating immigrants (but not emigrants) across the UK, down to the level of individual local authority areas. Therefore, the first step should be to assess how far these changes have improved the quality of these statistics. The best way of doing this is to compare – for the UK as a whole and for each local authority area – the results of the 2011 Census with the population estimates rolled forward from the 2001 Census. This should use both the original population estimates series and the one produced with the recent methodological improvements in order to see how much each of these deviates from the best estimate provided by the latest Census. The discrepancies, broken down by sex, age, country of birth and other personal characteristics, will provide pointers as to the source of any remaining problems and the actions needed to address them.

21. Even before the results of these checks are known, it 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.

22. The idea of using e-Borders is potentially a very attractive one, if it provides a full count of people entering and leaving the UK and also allows the matching of records over time in order to identify the actual length of time that individual people have stayed in or out of the UK. While this system would not provide the richness of information obtained via the IPS, it could be used in conjunction with the IPS to produce improved counts for the IPS-derived information to be grossed up to.

23. The other options relate to intensification of efforts to use other statistical sources to check and, where appropriate, amend the current methodology for estimating migration. This is particularly relevant for immigration counts because the people involved can be covered by surveys and administrative data sets. Besides using the decennial Census (see above), sample surveys like the APS provide information on country of birth and year of arrival in the UK on a more frequent basis. Similarly, administrative data sources such as those maintained by the NHS and central government departments can be used for this purpose. The experience of matching data sets being gained by the Beyond 2011 programme should be drawn upon here.

24. By contrast, emigration poses a stiffer challenge as its better estimation requires identifying people who are no longer living in the UK. Commonly, there is no incentive to deregister from administrative lists (e.g. NHS, NI) on departure, nor by definition will these people be included in surveys. Nevertheless, there are a number of indirect methods, including:

· surveys which ask about any of their household or family members that are living abroad (but this approach provides only a partial picture as it will miss cases where all members have moved);

· collection of data from non-UK statistical agencies on the numbers of people arriving in their countries from the UK (valuable in that, just as for the UK, most countries possess more accurate records on their immigrants than on their emigrants);

· examination of the UK’s three Longitudinal Studies to identify people that cannot be traced from one Census to the next and have not been registered as deaths (or possibly a separate customised analysis using higher sampling fractions); and

· exploration of ways of better documenting UK nationals living abroad through pensions and other administrative data (possibly as part of or a follow-on from Beyond 2011).

January 2013