Public Administration CommitteeCorrespondence between Bernard Jenkin MP, Chair of PASC, and Sir Andrew Dilnot CBE, Chair of the UK Statistics Authority (5STATS 14)


I am writing about the Government’s net migration target, following the recent oral evidence session we held on the topic of “migration statistics”, as part of our wider programme of work on statistics and their use in government.

As you are no doubt aware, the Government aims to reduce annual net inward migration to the UK “from the hundreds of thousands back down to the tens of thousands” by the end of the current Parliament. In the period 2005 to 2009, annual average net migration to the UK was around 200,000; so in order for the Government to achieve its target, the official ONS estimate of net migration would need to fall by around 100,000, which I understand is roughly half of annual average net migration in the period before the Government took office. I am told that net migration could be 35,000 higher or lower than the number given. Therefore, the true value of net migration falls within a range of around 70,000.

Given the size of the uncertainty surrounding the official estimate of net migration, relative to recent and targeted levels of net migration, is it, in your view, suitable for measuring progress against the Government’s net migration target?


Net Migration Statistics

Net migration is calculated as the, relatively small, difference between the much larger estimates for immigration and emigration. Each of these estimates is subject to considerable uncertainty in terms of any one measurement. I have included the latest figures1 below to illustrate this:

Year ending September 2013

Confidence interval



+/- 26,000



+/- 22,000

Net migration


+/- 34,000

ONS’s “confidence intervals” for net migration mean, in effect, that around 19 times out of 20 the true figure will lie within plus or minus 34,000 of the estimate—and one time in 20 it may be even further away. As you say in your letter, that is a broad range. However, the enclosed graph, prepared by the Statistics Authority on the basis of the official data, shows that over time a fairly clear picture emerges. The graph uses quarterly data (on a rolling average for four quarters basis) and runs up to mid-2012. We can have quite a lot of confidence that the picture that emerges is broadly right, even if the individual estimates at a point in time are more uncertain. The degree of uncertainty is illustrated by the short grey lines. What the graph shows is that net migration started to fall from about early 2011.

Nonetheless, commentators would be wise to treat the results for any one year or quarter with some caution for the reasons noted above. Looking back further, at the period between the 2001 and 2011 censuses, it is possible that both immigration and emigration were under-estimated to some degree due to the difficulty in obtaining reliable information from travellers at UK ports of entry. The 2011 Census data has proved to be an important check on any “drift” due to such problems and this is just one example of the way that Census results can “benchmark” other estimates.

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. The calculation of confidence intervals for estimates such as immigration and emigration is complex as they combine both survey data, from the International Passenger Survey, and administrative data, including asylum data. Calculating confidence intervals for the difference between these two estimates adds another layer of complexity, and I have asked ONS officials to consider whether there are statistical methods that could reduce the confidence intervals by considering trends over time.

The Statistics Authority will soon be publishing a report on the International Passenger Survey (IPS) which is the primary source for these estimates. That will draw attention to the weaknesses inherent in relying on a survey of passengers at UK ports of entry. The IPS is the best statistical source currently available. The UK does not have the sort of register of the population that some European countries have. Until more systematic information is collected from all passengers at ports of entry, statisticians will have to rely on the IPS. The Home Office’s e-Borders programme may offer some improvement in the longer term.

June 2013

1 Source:

Prepared 26th July 2013