Dear Chris

You contacted us about the status of ONS’s migration estimates by quarter in November, and we discussed the issue in early February. This letter sets out my conclusions following our discussions.

The background is that ONS published data on migration by quarter[1] in response to user requests. In publishing the data, ONS stated clearly that these data should not be regarded as official statistics.

My conclusions are based on the analysis conducted by my team, included as an Annex to this letter. In summary:

  • Whether the data by quarter are classed as official statistics is a matter of judgement for the producer of the statistics. ONS has considered the National Statistician’s guidance[2] on this matter in order to form its view.
  • In reviewing the ONS’s judgement we considered each of the areas set out in the National Statistician’s guidance: materiality, public interest, regularity, methodology, coverage and publication. We consider that data by quarter meet these criteria with the exception of methodology and coverage.
  • The data by quarter are based on a smaller sample and the complete methodology[3] cannot be applied. As explained in the Annex, for these reasons we consider that ONS’s approach is justified.
  • We consider that the intention to not class these as official statistics was to provide a clear indication to users of the associated quality issues and the limitations of the data by quarter.
  • ONS has provided some commentary to users to support user interpretation, but a lot of weight is being placed on user understanding of the term “official statistics”, which may not be familiar to all users.

I therefore consider that the ONS’s decision not to classify as official statistics is defensible by reference to the relevant guidance. But it is not the only or best way to signal to users that they should treat the data by quarter with caution. We note that ONS did provide some guidance to users, but consider that ONS could have done more to support understanding of these data. We have asked ONS to:

  • provide guidance to users on the quality of the estimates and how they should interpret the data. These data are of high public interest and some users are likely to pay close attention to them, so fuller guidance on what can and cannot be inferred from them is important;
  • be clearer about how it reached its judgement that the estimates should not be published as official statistics;
  • more clearly demonstrate how it is applying the Code of Practice for Statistics in the production of the data by quarter. A statement of compliance with the Code would demonstrate which aspects of Trustworthiness, Quality and Value have been met and which have not, and therefore provide greater assurance around the production of these estimates and associated limitations.

ONS has sought to respond to these expectations in its February Migration Quarterly Statistics report[4], by including clearer information on how they reached the decision on the official statistics status of these estimates and updating the commentary on how the estimates should or should not be used.

These quarterly data are of interest because of the public interest in migration more broadly. The robustness and value of migration statistics remain a key area of focus for OSR. Our focus on migration statistics is reflected in our recent work:

  • Measuring migration using the International Passenger Survey has well known limitations, some of which we highlighted in our report on student migration statistics[5] in July 2017, which effectively led to the removal of the National Statistics designation from the student emigration statistics. ONS no longer publishes a headline figure for net student migration.
  • We wrote to the National Statistician[6] in March 2016 highlighting the importance of explaining the differences between estimates of migration produced by ONS and administrative data on the issuance of National Insurance numbers to non-UK individuals.
  • In October 2018, we published a review[7] of the way ONS handled changes to the data collection approach for the International Passenger Survey.

ONS and other migration statistics producers are currently working together on a planned transformation programme[8] which involves drawing on a wider range of government administrative sources beyond the International Passenger Survey to improve the quality and value of migration estimates. On 30 January ONS published an update on progress on the transformation journey[9].

We welcome this update from ONS. It will be essential to keep users engaged on developments. We will continue to monitor progress on migration statistics as part of our work programme for 2019-20, including reviewing whether the ambitions set out in the transformation programme are being met in an effective and timely way and are supporting users of the statistics.

I am copying this letter to Iain Bell, Deputy National Statistician with responsibility for migration statistics.

Yours sincerely,

Ed Humpherson
Director General for Regulation

[1] datasets/estimatesoflongterminternationalmigrationbyquarterderivedfromtheinternationalpassengersurvey


[3] methodologies/longterminternationalmigrationestimatesmethodology






[9] aresearchengagementreport