Migration Statistics Quarterly Report - information for users

This user information article provides more details of the quality and data sources used in the Migration Statistics Quarterly Report (MSQR). This includes:

• guidance on how to interpret confidence intervals (Section 2)
• guidance on the difference between provisional and final data (Section 3)
• guidance on interpretation and comparability of the data sources (Section 4 and 5)
• information on the nature and methodology of each source, highlighting any definitions or quality issues to be aware of when interpreting the data (Section 5)

Sample surveys are used when we would like to know something about a population of individuals but asking all of them is impractical. For example, we might wish to know how many people are migrating into and out of the UK each year. One way to find this out would be to ask everyone crossing the UK border about their migration intentions. However, this would be a very expensive and impractical approach. Instead a more cost-effective approach is to use a sample survey.

The International Passenger Survey (IPS) interviews a sample of passengers (around 700,000 to 800,000 per year) passing through entry and exit points to and from the UK. This sample is used, among other things, to provide data to estimate the true number of people migrating into and out of the UK.

As with all sample surveys, the estimates produced are based on only one of a number of possible samples of passengers that could have been drawn at a given point in time. Each of these possible samples would produce an estimated number of migrants, which may be different to the true value that would have been obtained if everyone passing through were interviewed.

2.1 How do we know how accurate our estimate is?

The different possible samples of passengers that could have been selected can be used to produce a sampling distribution for the figure we are trying to estimate. For example, if we are estimating immigration within a particular year, one sample may produce an estimate of 500,000, another may have resulted in an estimate of 515,000 and another may have produced an estimate of 490,000. If we could take a lot of samples like this and plot the estimates from each sample we would produce a chart of the sampling distribution of our estimate.

Assuming that the estimation method we use to produce the estimate for each sample is unbiased, the shape of the plot would follow the widely recognised normal distribution, where the most likely estimates of the true value are centred towards the middle and the least likely estimates are at the “tail ends”.

Figure 1: Normal distribution curve

Download this image Figure 1: Normal distribution curve

.png (10.4 kB)

In practice, in order to estimate the true value for a specific population of, say, the number of immigrants, we take one sample and produce a single estimate. We assume that the sampling distribution of our estimate would follow approximately a normal distribution, centred on the true value, and we can use a statistical formula to calculate the standard error (SE) around the estimate. This is a measure of the accuracy of the estimate.

As illustrated in Figure 1, 95% of the estimates would lie within 1.96 multiplied by the true standard deviation of the sampling distribution. This also works the other way round, so we can say that for 95% of random samples taken, our estimate will be no more than 1.96 multiplied by the standard error of that estimate away from the true value that we are trying to estimate. Using this knowledge, we can calculate a confidence interval around our estimate.

Confidence intervals are indicators of the extent to which the estimate may differ from the true value. The larger the confidence interval, the less precise is the estimate. The central value within the confidence interval is the best estimate of the true value. The confidence interval around the estimate captures the uncertainty of the estimate, and gives an interval within which we can say that there is a high probability that the true value lies.

At the 95% confidence level, over many repeats of the sample under the same conditions, we would expect the confidence interval to contain the true value 95 times out of 100. Equivalently, we can say that there would be a 1 in 20 chance that the true value would lie outside of the range of the 95% confidence interval.

2.2 Interpreting change

Changes in the estimates from the IPS from one period to the next may occur simply by chance. In other words, the change may be due to which individuals were selected to answer the survey, and may not represent any real-world change in migration. We are able to measure whether this is likely to be the case using standard statistical tests at the 5% level of “statistical significance”. When we report on statistical significance, we provide an assessment of how likely it is that we would see results as unusual as these if the true value of the population remained unchanged. The phrase "statistically significant at the 5% level" indicates that, if the true value of the population remained unchanged, a result like this would occur less than 5% of the time.

You will observe that the confidence intervals become larger for more detailed estimates (such as citizenship by reason for migrating). This is because the number of people in the sample who have these specific characteristics (such as EU8 citizens arriving to study in the UK) is smaller than the number of people sampled within a category at a higher level (such as the number of total EU citizens arriving to study in the UK). Therefore users of migration statistics are advised to use the highest level breakdown of data where possible.

When using estimates of detailed characteristics, you are advised to be cautious about drawing conclusions regarding changes in figures from one time period to the next. To assist with this, shading has been applied to cells within the published tables to indicate where a change is statistically significant from the previous year. This means that the change is likely to reflect a real change in migration patterns rather than occurring as a result of the sampled passengers being different in characteristics by chance.

Long-Term International Migration (LTIM) estimates are calculated from IPS estimates with additional data to account for flows of asylum seekers, non-asylum enforced removals, refugees resettled under the various resettlement schemes, flows from Northern Ireland and visitor and migrant switchers. LTIM estimates are shown with the confidence intervals for the IPS component of the estimate in order to give users an indication of the accuracy of the estimate. Similarly the uncertainty associated with the IPS component of the estimate is used to calculate statistically significant changes in the LTIM estimates. However, when interpreting these confidence intervals and statistically significant changes, you should be aware that there is no method for quantifying the error associated with the non-survey components of LTIM, which is unlikely to be random.

2.3 Revisions to net migration estimates

In April 2014, we published a report on the quality of LTIM from 2001 to 2011. This review found that prior to improvements on the IPS in 2009, net international migration was likely to have been underestimated during the middle part of the decade. Net migration estimates for 2001 to 2011 have been revised, and a guidance note for net international migration revisions to users on interpreting these revised estimates alongside published LTIM estimates has been published.

The estimates presented by us contain final Long-Term International Migration (LTIM) and International Passenger Survey (IPS) data for all quarters to December 2016 and provisional data for March and June 2017. The final LTIM data for 2016 were released in November 2017. Final LTIM data for 2017 will be released in November 2018.

The estimates presented by us contain final International Passenger Survey (IPS) estimates for all rolling years to year ending (YE) December 2016 and provisional estimates for YE March 2017.

The full set of final LTIM and IPS tables for 2015 were released in December 2016. Final LTIM and IPS tables for 2016 will be released in November 2017.

Provisional figures allow for a timely comparison of recent migration patterns on a quarterly basis. However, these are subject to change as their calculation is based upon provisional data. The final LTIM estimates are considered to provide a more reliable picture of migration and allow for annual comparisons over time.

Estimates of LTIM are about 90% based on data from the IPS, supplemented by adjustments made for asylum seekers, non-asylum enforced removals, flows from Northern Ireland and visitor and migrant switchers. Therefore, the majority of the differences between provisional and final LTIM estimates can be accounted for by the replacement of provisional IPS data with final IPS data. Nonetheless, the differences between overall provisional and final estimates of long-term international migration from the IPS are minimal. Table 1 shows the specific difference between provisional and final IPS data for years ending March, June, September and December 2016.

Once the final IPS data are incorporated in the LTIM estimates, final data for asylum seeker flows, non-asylum enforced removals and flows from Northern Ireland are incorporated to produce the final LTIM estimates. These adjustments are typically negligible. For example, Table 2 shows that differences are negligible between the provisional LTIM figures (including final IPS data) for year ending December 2016 published in August 2017, and the final LTIM figures for the same period published in November 2017.

There are a range of different data sources used in the Migration Statistics Quarterly Report (MSQR) and it is important to understand what each source relates to and the ways in which it may be interpreted.

4.1 Classes of migration information

There are three broad classes of international migration information in the MSQR: flows, registrations and stocks.

Flows

These are direct measures of migration – that is how many people are moving from one place to another. Information on international flows is available in two Office for National Statistics (ONS) datasets: Long-Term International Migration (LTIM) and International Passenger Survey (IPS) estimates of long-term international migration. IPS data are available in further detail than LTIM data. Note that both of these datasets only consider migrants who stated that they were intending to stay in or be absent from the UK for a period of at least 12 months.

Registrations

These are measures of the number of people applying for and/or being granted permission to work or stay in the UK, including Department for Work and Pensions (DWP) information on National Insurance number (NINo) allocations to adult overseas nationals, and the Home Office’s data on visas, asylum and grants of settlement.

Stocks

These are measures of the numbers of people resident in the UK. Stocks information is available in ONS’s and National Records of Scotland’s (NRS’s) data on population by country of birth and nationality. Stocks data are published annually in August.

4.2 Comparing International Passenger Survey (IPS) estimates with Long-Term International Migration (LTIM) estimates

The provisional quarterly LTIM estimates, which were introduced to the MSQR in August 2010, can be compared directly with the annual LTIM estimates published each November. For data published from February 2012 onwards, provisional LTIM figures are available in two tables for citizenship and main reason for migration.

The most timely cross-tabulated data are still the provisional IPS estimates. The IPS table “citizenship by main reason for migration” provides valuable information about migration patterns into and out of the UK, but this level of detail is not available for the additional adjustments that are applied to LTIM estimates. Particular care is needed when comparing these with LTIM.

The IPS is the main component of LTIM but does not take into account adjustments for:

• people whose length of stay changes from their original intentions
• additional information on international migration to and from Northern Ireland
• asylum seekers, non-asylum enforced removals and refugees resettled under the various resettlement schemes

The IPS is supplemented by these adjustments, resulting in LTIM estimates.

4.3 Comparing flows, registrations and stocks data

Flows measure all people who state their intention to move for 12 months or more. However, stocks are based on the resident population and are measured from the Annual Population Survey (APS), which covers households only and not residents living in communal establishments.

Data on flows, registrations and stocks may display apparently contradictory trends, but are valuable in their own right, as explored in more detail within Comparing sources of international migration: August 2017.

For example, it may be that the LTIM data indicate steady numbers of immigrants, but the NINo statistics show an increased number of successful registrations. However, the LTIM data describe people stating an intention to stay at least a year, whereas the NINo statistics include people who came to work in the UK for a short period of time (for example, for a summer job) and then left again. In addition, NINo allocations do not necessarily reflect a recent move to the UK at all, as an overseas national may already have been in the UK for several years (for example, as a student or a dependant) before they decide to seek employment.

In May 2016, ONS published a Note on the difference between National Insurance Number registrations and the estimate of Long-Term International Migration.

Alternatively, the LTIM inflows and registrations data may suggest downward trends while the stock of people born outside the UK has gone up. This can be explained by a situation as follows. Assume a starting point of 0 people from a particular country of birth in the UK. If 100,000 such people enter the UK in the first year (and all stay) but in the next year the inflow halves so only 50,000 enter (and all stay), the total stock would have risen to 150,000.

In other words, stocks of people born outside the UK can continue to rise despite reduced inflows, and only go down if the combined totals of emigration and deaths exceed immigration.

Again, this is outlined in more detail in a August 2017 article: Comparing sources of international migration.

4.4 Stocks data by country of birth and nationality

The population by country of birth and nationality data have a link with migration but are only partly influenced by current migration trends.

People born overseas include recent migrants but also people who have lived in the UK for decades, many of whom now have British nationality. There are also people who were born overseas but have always been British nationals – for example, the children of armed forces personnel who were living abroad at the time of the birth.

Similarly, people of non-British nationality may have been in the UK for decades, or may have been born in the UK to non-British parents and have always lived in the UK.

Therefore just because someone was born abroad or is not of British nationality does not necessarily mean they are what many people would consider to be an “immigrant”.

Further information on interpretation and comparison of international migration flows and stocks data sources is available in the following article: Note on the differences between Long-Term International Migration flows derived from the International Passenger Survey, and estimates of the population obtained from the Annual Population Survey.

Our quarterly publication, the Migration Statistics Quarterly Report (MSQR), describes a range of data sources, which reflect different aspects of migration.

The different sources (Figure 1) provide a useful comparison for Long-Term International Migration (LTIM), however, they cover different groups of people and so differences between them are to be expected. This Venn diagram shows the main similarities and differences between the three data sources.

There are exceptions to these main differences, concerning mainly small groups of migrants. These are detailed in Table 3.

The Venn diagram compares sources within the 12-month reference period. When comparing LTIM with other sources it must be noted that, particularly with National Insurance numbers (NINos), migrants may not be captured in the comparison source in the LTIM reference period.

For example, a migrant may have arrived in September 2013 and been counted as a migrant by LTIM. They register for a NINo in November 2014, and so were not counted in NINo registrations until the following year.

In May 2016, Office for National Statistics (ONS) published a Note on the difference between National Insurance Number registrations and the estimate of Long-Term International Migration.

Figure 2: Comparison of coverage of MSQR data sources

Notes:

1. “European” includes migrants from European Economic Area (EEA) countries and Switzerland.

2. * LTIM records main reason for migrating but not all reasons, for example someone arriving as a student could go on to apply for a NINo in order to do some part-time work. In this example, LTIM would record a student and the student would also be counted in NINo registrations.

3. ** Visas excluding visas for visit or transit.

Download this image Figure 2: Comparison of coverage of MSQR data sources

.png (129.5 kB)

This table shows which data sources cover which group of migrants.

Table 4: Summary of datasets used in Migration Statistics Quarterly Report

A guide to Home Office administrative datasets and their coverage is given in the User Guide to Home Office Immigration Statistics.

Last updated November 2017. Produced in partnership with the Home Office and the Department for Work and Pensions (DWP).